TFL1-Like Proteins in Rice Antagonize Rice FT-Like Protein in Inflorescence Development by Competition for Complex Formation with 14-3-3 and FD

TFL1-Like Proteins in Rice Antagonize Rice FT-Like Protein in Inflorescence Development by... Abstract Hd3a, a rice homolog of FLOWERING LOCUS T (FT), is a florigen that induces flowering. Hd3a forms a ternary ‘florigen activation complex’ (FAC) with 14-3-3 protein and OsFD1 transcription factor, a rice homolog of FD that induces transcription of OsMADS15, a rice homolog of APETALA1 (AP1), which leads to flowering. TERMINAL FLOWER 1 (TFL1) represses flowering and controls inflorescence architecture. However, the molecular basis for floral repression by TFL1 remains poorly understood. Here we show that RICE CENTRORADIALIS (RCN), rice TFL1-like proteins, compete with Hd3a for 14-3-3 binding. All four RCN genes are predominantly expressed in the vasculature, and RCN proteins are transported to the shoot apex to antagonize florigen activity and regulate inflorescence development. The antagonistic function of RCN to Hd3a is dependent on its 14-3-3 binding activity. Our results suggest a molecular basis for regulation of the balance between florigen FT and anti-florigen TFL1. Introduction Timing of flowering and inflorescence architecture are important for plant reproductive success and are controlled by both environmental and endogenous conditions. One of the most important environmental cues for flowering is photoperiod. Photoperiodic flowering is regulated by the interplay between light signaling and a circadian clock pathway (Song et al. 2015). In Arabidopsis, GIGANTEA (GI), a circadian clock-regulated nuclear protein, up-regulates CONSTANS (CO) (Mizoguchi et al. 2005, Sawa et al. 2007). CO is a CCT-domain transcription factor and activates the expression of FLOWERING LOCUS T (FT) (Suárez-López et al. 2001). FT and its homologs in other plant species act as the key protein component of florigen, a systemic floral inducer (Kardailsky et al. 1999, Kobayashi et al. 1999, Corbesier et al. 2007, Jaeger and Wigge 2007, Lin et al. 2007, Mathieu et al. 2007, Tamaki et al. 2007). The GI–CO–FT module for photoperiodic flowering is highly conserved in rice, a short-day (SD) plant. In rice, OsGI (rice ortholog of GI), Heading date 1 (Hd1; rice ortholog of CO) and Heading date 3a (Hd3a; a rice FT) make up the GI–CO–FT module (Kojima et al. 2002, Hayama et al. 2003, Tsuji et al. 2013). The dual function of Hd1 represents a major difference between Arabidopsis, a long-day (LD) plant, and rice; Hd1 activates and represses the expression of Hd3a under SD and LD conditions, respectively (Hayama et al. 2003, Ishikawa et al. 2011). FT encodes an approximately 20 kDa globular protein similar to phosphatidylethanolamine-binding protein (PEBP) (Kardailsky et al. 1999, Kobayashi et al. 1999). Loss of function of FT delays flowering and its ectopic expression causes early flowering and a terminal flower phenotype (Kardailsky et al. 1999, Kobayashi et al. 1999). FT protein is produced in the leaves under inductive day length conditions and transported through the phloem to the shoot apical meristem (SAM) to change the growth phase from vegetative to reproductive (Corbesier et al. 2007, Tamaki et al. 2007, Tamaki et al. 2015). The floral promotion activity of FT requires interaction with FD, a basic/leucine zipper (bZIP) transcription factor (Abe et al. 2005, Wigge et al. 2005). Phosphorylation of the conserved C-terminal motif of FD (SAP motif) is essential for interaction with FT and activation of APETALA1 (AP1), a floral meristem identity gene (Abe et al. 2005, Taoka et al. 2011, Kawamoto et al. 2015). In rice, it has been shown that the FT–FD interaction is mediated by 14-3-3 (Taoka et al. 2011). 14-3-3 is a highly conserved scaffold protein that bridges two proteins through phosphorylated 14-3-3-binding motifs and is involved in a variety of plant regulatory pathways (de Boer et al. 2013, Ormancey et al. 2017). Hd3a interacts with 14-3-3 in the cytoplasm and then the Hd3a–14-3-3 complex interacts with OsFD1, a rice FD homolog, in the nucleus through the phosphorylated SAP motif of OsFD1 (Taoka et al. 2011). Analysis of the crystal structure revealed that Hd3a makes a heterohexameric protein complex of two Hd3a, dimeric 14-3-3 and dimeric OsFD1, which is called the florigen activation complex (FAC) (Taoka et al. 2011). The amino acid residues critical for the interaction between FAC components are highly conserved among angiosperms, and mutations that impair the interaction reduce the activation of OsMADS15, a rice AP1 homolog (Taoka et al. 2011). TERMINAL FLOWER 1 (TFL1) is known to counteract and fine-tune florigen activity (Kobayashi et al. 1999, Shalit et al. 2009, Lifschitz et al. 2014). Loss of function of TFL1 causes early flowering and a terminal flower phenotype in Arabidopsis (Bradley et al. 1997). TFL1 also encodes a PEBP-like protein with similarity to FT (Bradley et al. 1997, Kardailsky et al. 1999, Kobayashi et al. 1999). However, small differences between TFL1 and FT in distribution of charged amino acid residues, structure of putative anion-binding pockets and the amino acid sequence of the external loop confer antagonistic activity to TFL1 (Hanzawa et al. 2005, Ahn et al. 2006, Ho and Weigel 2014). In the mature SAM, TFL1 mRNA is restricted to the center of the SAM, but TFL1 protein moves from cell to cell and co-ordinates the cell identity (Conti and Bradley 2007). In contrast, Arabidopsis thaliana CENTRORADIALIS homolog (ATC), another TFL1 in Arabidopsis, is expressed in the vasculature under SD conditions and is translocated to repress flowering (Huang et al. 2012). The graft-transmissible activity of floral inhibition suggests that ATC may be a systemic floral inhibitor, an anti-florigen that was proposed in tobacco (Lang et al. 1977, Huang et al. 2012). In rice, the four isoforms of RICE CENTRORADIALIS (RCN) are classified ino the TFL1 subfamily (Nakagawa et al. 2002, Liu et al. 2013). Overexpression of RCN1 and RCN2 delays flowering and increases the number of branches (Nakagawa et al. 2002). RCN-knockdown plants exhibit small panicles with reduced branches (Liu et al. 2013). TFL1 and TFL1-like proteins in several plant species are also known to interact with FD (Abe et al. 2005, Wigge et al. 2005, Hanano and Goto 2011, Huang et al. 2012, Higuchi et al. 2013, Varkonyi-Gasic et al. 2013, Ho and Weigel 2014, Ryu et al. 2014). The interaction with FD and conserved putative 14-3-3-binding sites in the TFL1 family suggest that TFL1 also makes a FAC-like complex to repress FT (Tsuji et al. 2013). However, the involvement of 14-3-3 in the antagonistic role of TFL1 has not been analyzed yet. In this study, we examined the spatiotemporal expression of RCN and the protein interaction of RCN. We show that RCN protein is transported to the SAM and can interact with OsFD1 to repress florigen activity in a 14-3-3-dependent manner. It suggests that competitive binding of RCN and Hd3a to 14-3-3 plays an important role in ensuring the optimal reproductive development of rice. Results Tissue-specific expression of RCN genes Phylogenetic tree analysis of 19 PEBP-like genes in rice identified four TFL1-like genes, RCN1–RCN4 (Supplementary Fig. S1). To examine the spatio-temporal expression of RCN RNA, semi-quantitative reverse transcription–PCR (RT–PCR) analysis was performed. RCN RNA was strongly detected in stem and root, from vegetative to late reproductive phase (Fig. 1A;Supplementary Fig. S2). In contrast, Hd3a is strongly expressed in the leaf under floral inductive SD conditions (Kojima et al. 2002, Tamaki et al. 2007). Under non-floral inductive LD conditions, RCN expression was not significantly different from that under SDs (Supplementary Fig. S2). In the SAM, no RCN RNA was detected at any growth phase (Fig. 1A). For a more sensitive assay, RT–PCR samples shown in Fig. 1A were subjected to a second round of PCR amplification, and consistently no amplification products were detected in the SAM (Supplementary Fig. S2A). To examine the expression pattern at the cellular level, the GUS (β-glucuronidase) reporter was fused with 5' upstream (approximately 3 kb) and coding regions of RCN genes (Supplementary Fig. S3A) and histochemical analysis of the transgenic rice plants with these chimeric genes (gRCN-GUS) was performed. GUS signal from gRCN1-GUS plants was detected in the vascular tissue of the stem and root (Fig. 1B;Supplementary Fig. S3). In the cross-section of the stem, GUS signal was localized in the phloem of the vasculature (Fig. 1B). Around the shoot apex, GUS signal was detected in the presumptive vascular tissue and beneath the SAM throughout the vegetative to reproductive phase (Fig. 1B). However, no GUS signal was detected in the SAM. Similar GUS staining patterns were observed in gRCN2-GUS and gRCN4-GUS plants (Fig. 1B;Supplementary Fig. S3). In gRCN3-GUS plants, GUS signal was also detected in the phloem, while it was not detected around the shoot apex (Fig. 1B). In all gRCN-GUS transgenic plants examined, no detectable GUS signal was found in the SAM from the vegetative to reproductive phase. Fig. 1 View largeDownload slide Expression of RCN in rice. (A) Expression of RCN RNA in rice. RCN expression was analyzed by semi-quantitative RT–PCR at the vegetative (V), transition (T), reproductive (R) and late reproductive (LR) phase. Ubiquitin (Ubiq) RNA was analyzed as a control. LB, leaf blade; LS, leaf sheath; ST, stem; RO, root; SAM, shoot apical meristem. (B) GUS staining images of gRCN-GUS transgenic rice plants. SAM at the vegetative (V) and reproductive (R) phase and the vasculature of the stem are shown. Asterisks indicate the SAM. A white triangle indicates GUS signal derived from the leaf primordium. Red triangles indicate phloem. Scale bar = 50 μm in the SAM and 20 μm in the vasculature. (C) Localization of RCN–GFP protein in gRCN-GFP transgenic rice plants. Confocal images of the SAM from the vegetative (V) to reproductive (R) phase and the stem vasculature in gRCN-GFP transgenic rice plants are shown. IM, inflorescence meristem; PBM, primary branch meristem; BR, bract. Scale bar = 50 μm. Fig. 1 View largeDownload slide Expression of RCN in rice. (A) Expression of RCN RNA in rice. RCN expression was analyzed by semi-quantitative RT–PCR at the vegetative (V), transition (T), reproductive (R) and late reproductive (LR) phase. Ubiquitin (Ubiq) RNA was analyzed as a control. LB, leaf blade; LS, leaf sheath; ST, stem; RO, root; SAM, shoot apical meristem. (B) GUS staining images of gRCN-GUS transgenic rice plants. SAM at the vegetative (V) and reproductive (R) phase and the vasculature of the stem are shown. Asterisks indicate the SAM. A white triangle indicates GUS signal derived from the leaf primordium. Red triangles indicate phloem. Scale bar = 50 μm in the SAM and 20 μm in the vasculature. (C) Localization of RCN–GFP protein in gRCN-GFP transgenic rice plants. Confocal images of the SAM from the vegetative (V) to reproductive (R) phase and the stem vasculature in gRCN-GFP transgenic rice plants are shown. IM, inflorescence meristem; PBM, primary branch meristem; BR, bract. Scale bar = 50 μm. RCN proteins are transported to the SAM Our RT–PCR and promoter–GUS analysis strongly suggested that no RCN RNA is expressed in the SAM. However, it has been reported that RCN regulates the floral transition and inflorescence development (Nakagawa et al. 2002, Liu et al. 2013, Wang et al. 2015). Therefore, we examined the localization of RCN proteins in plants. The same 5' upstream and coding regions of RCN genes as used in the GUS analysis were fused with the green fluorescent protein (GFP) coding region (gRCN-GFP) and fluorescence from RCN–GFP protein in the transgenic rice plants was observed with confocal microscopy (Fig. 1C;Supplementary Fig. S4). Consistent with the semi-quantitative RT–PCR and GUS staining, a strong RCN–GFP signal was observed in the stem vasculature for all four RCN isoforms (Fig. 1C). In addition, RCN–GFP signal was also observed in the SAM, from the vegetative to reproductive phase (Fig. 1C), indicating the transport activity of RCN proteins from the vascular tissue to the SAM. At the late reproductive phase, localization of RCN–GFP in the shoot apex was confined to the floral meristem for RCN1, RCN2 and RCN3, whereas RCN4–GFP was observed in and beneath the floral meristem (Supplementary Fig. S4). Translocation ability of RCN3–GFP to the SAM was also observed when RCN3–GFP was expressed under the control of a phloem-specific promoter from Agrobacterium rhizogenes, prolC (Supplementary Fig. S5). These results suggested that RCN proteins are predominantly synthesized in the phloem and translocated to the SAM. RCN can interact with 14-3-3 and OsFD1 First we tested the interaction of RCN with 14-3-3, because Hd3a interacts with OsFD1, a rice FD homolog, through 14-3-3 binding (Taoka et al. 2011). The critical amino acid residues of Hd3a for 14-3-3 binding were completely conserved among all four RCN isoforms (Fig. 2A). In a yeast two-hybrid assay, the four RCNs interacted with GF14b, a rice 14-3-3 (Fig. 2B). The 14-3-3 isoform preference of RCN was also comparable with that of Hd3a; the four RCNs interacted with GF14b, GF14c, GF14e and GF14f (Supplementary Fig. S6B). The amino acid substitution in RCN corresponding to impaired interaction between Hd3a and 14-3-3 reduced the interaction between RCN and 14-3-3 (Fig. 2B). The direct protein–protein interaction was tested by glutathione S-transferase (GST) pull-down assays using highly purified recombinant proteins. GST–GF14c pulled down the four recombinant RCN proteins with comparable efficiency (Fig. 2C). OsFD1 peptide was not pulled down with GST–RCN3, whereas GST–GF14c co-precipitated both RCN3 and OsFD1 (Fig. 2D). As with Hd3a, RCN3 was pulled down with GST–OsFD1 only in the presence of GF14c (Fig. 2D). The four RCNs interacted with OsFD1 in yeast (Supplementary Fig. S7B). Consistent with previous reports that phosphorylation of serine/threonine in the C-terminal SAP motif of Arabidopsis FD and rice OsFD1 is essential for 14-3-3 binding (Taoka et al. 2011, Kawamoto et al. 2015), alanine substitution of Ser192 (S192A) of OsFD1 prevented interaction with RCN, but phosphomimic glutamate substitution (S192E) retained it (Supplementary Fig. S6A). Endogenous yeast 14-3-3 was hypothesized to mediate the interaction between Hd3a and OsFD1, because Hd3a did not interact directly with OsFD1 in vitro (Taoka et al. 2011). In support of this idea, the two yeast 14-3-3s, brain modulosignalin homolog 1 (BMH1) and BMH2, interacted with RCN and Hd3a (Supplementary Fig. S6C) and OsFD1 (Supplementary Fig. S6D). These interactions were prevented with m2 substitution in the conserved 14-3-3 binding residues in RCN3 (R62K/R130K) and in Hd3a (R64K/R132K) or S192A substitution in OsFD1 (Supplementary Fig. S6). The RCN3–GF14b–OsFD1 interaction was also tested in rice protoplasts. Fluorescence from RCN3 tagged with mCherry was detected in both the nucleus and cytoplasm (Fig. 2E). In a bimolecular fluorescence complementation (BiFC) assay, fluorescence from RCN3–GF14b interaction was detected in the cytoplasm, and the fluorescence was significantly reduced by m2 substitution in RCN3 (Fig.2F;Supplementary Fig. S7D). BiFC fluorescence from RCN3–OsFD1 interaction was observed predominantly in the nucleus and m2 substitution also significantly reduced the interaction with OsFD1 (Fig. 2G;Supplementary Fig. S7E). The subcellular localization of RCN3 was not affected by m2 substitution (Supplementary Fig. S7A). These localization and interaction features were similar to those of RCN1 (Supplementary Fig. S7) and Hd3a (Taoka et al. 2011), suggesting that RCN interacts with OsFD1 through 14-3-3 binding, like Hd3a. Fig. 2 View largeDownload slide Interaction among RCN, 14-3-3 and OsFD1. (A) Schematic representation of RCN. The critical amino acid residues for Hd3a–14-3-3 interaction (R64, P96, F103 and R132) and their corresponding residues in RCN are shown. (B) Yeast two-hybrid assay between RCN and 14-3-3. 3-AT, 3-aminotriazole; His, histidine. (C) GST pull-down assay using GST–GF14c and RCN. The signal intensity relative to that of co-precipitated RCN3 is shown below. (D) GST pull-down assays using (from left to right) GST–RCN3 and OsFD1 peptide; GST–GF14c and RCN3; GST–GF14c and OsFD1; and GST–GF14c and RCN3 with or without GF14c. (E) Confocal images of cells expressing RCN3–mCherry. NLS–CFP was co-expressed as a nuclear marker protein. (F) BiFC assays showing interactions of RCN3–GF14b (left) and RCN3 m2–GF14b (right). Venus fluorescence in cells expressing the indicated proteins tagged with the N- or C-terminal halves of Venus is shown. mCherry was co-expressed as a marker for transformation and for nuclear and cytoplasmic localization. RCN3 m2: RCN3 R62K/R130K. Scale bar = 5 μm. (G) BiFC assays showing interactions of RCN3–OsFD1 (left) and RCN3 m2–OsFD1 (right). Scale bar = 5 μm. Fig. 2 View largeDownload slide Interaction among RCN, 14-3-3 and OsFD1. (A) Schematic representation of RCN. The critical amino acid residues for Hd3a–14-3-3 interaction (R64, P96, F103 and R132) and their corresponding residues in RCN are shown. (B) Yeast two-hybrid assay between RCN and 14-3-3. 3-AT, 3-aminotriazole; His, histidine. (C) GST pull-down assay using GST–GF14c and RCN. The signal intensity relative to that of co-precipitated RCN3 is shown below. (D) GST pull-down assays using (from left to right) GST–RCN3 and OsFD1 peptide; GST–GF14c and RCN3; GST–GF14c and OsFD1; and GST–GF14c and RCN3 with or without GF14c. (E) Confocal images of cells expressing RCN3–mCherry. NLS–CFP was co-expressed as a nuclear marker protein. (F) BiFC assays showing interactions of RCN3–GF14b (left) and RCN3 m2–GF14b (right). Venus fluorescence in cells expressing the indicated proteins tagged with the N- or C-terminal halves of Venus is shown. mCherry was co-expressed as a marker for transformation and for nuclear and cytoplasmic localization. RCN3 m2: RCN3 R62K/R130K. Scale bar = 5 μm. (G) BiFC assays showing interactions of RCN3–OsFD1 (left) and RCN3 m2–OsFD1 (right). Scale bar = 5 μm. RCN competes with Hd3a for 14-3-3 binding Interaction of RCN with 14-3-3 and OsFD1 led to a hypothesis that RCN antagonizes Hd3a through competition with 14-3-3 binding in formation of the FAC. To test this hypothesis, we first examined the relationship between floral repression and 14-3-3 binding activity of RCN. OsMADS15 is a floral meristem identity gene in rice that regulates the floral transition (Kobayashi et al. 2012) and is activated by Hd3a (Komiya et al. 2008). In rice protoplasts, co-expression of Hd3a and OsFD1 can activate the expression of OsMADS15 (Fig. 3A). However, the activation was significantly compromised when RCN was co-expressed, and the reduction of OsMADS15 was not observed when the RCN3 m2 mutant was co-expressed (Fig. 3A). This suggests that 14-3-3 binding of RCN plays an important role in acting against Hd3a. The competition between Hd3a and RCN for 14-3-3 binding was also analyzed by in vitro competitive pull-down assay with highly purified recombinant proteins. Various ratios of Hd3a–hemagglutinin (HA) to RCN3 protein were mixed, and Hd3a-HA and RCN were pulled down with GST–GF14c (Fig. 3B). The 50% inhibitory concentration (IC50) values were calculated from the signal intensity of the precipitated proteins and evaluated as 0.6 μM for RCN3 against Hd3a–GF14c binding and as 3.0 μM for Hd3a against RCN3–GF14c binding (Supplementary Fig. S8), indicating that RCN3 can bind to GF14c with approximately five times higher affinity than Hd3a in vitro. To assess if RCN can competitively form an FAC-like complex, we performed BiFC analyses between Hd3a and RCN. When RCN3 proteins fused with the N- and C-terminal half of mVenus were co-expressed with OsFD1 in rice protoplasts, fluorescence from RCN3-RCN3 interaction was observed in the nucleus (Supplementary Fig. S9A). The BiFC signal was significantly reduced when the m2 substitution was introduced in RCN3 (Supplementary Fig. S9A, B). The 14-3-3-dependent BiFC interaction of RCN3-RCN3 is consistent with a proposed structure of the FAC, a hexameric protein complex comprised of two Hd3a, a 14-3-3 dimer and an OsFD1 dimer (Taoka et al. 2011). When BiFC interaction of Hd3a-Hd3a was tested in the same way, 14-3-3-dependent BiFC interaction was observed (Supplementary Fig. S9A, B), supporting the notion that Hd3a makes a FAC. If RCN competes with 14-3-3 binding to Hd3a for FAC formation, a heteromeric FAC-like complex of Hd3a and RCN could be formed. As expected, 14-3-3-dependent BiFC interaction between Hd3a and RCN3/RCN1 was observed (Fig. 3C;Supplementary Fig. S9A, B), suggesting that Hd3a and RCN3 are exchangeable in formation of a FAC-like complex. Recently, direct interaction between FT isosforms was reported to fine-tune the flowering time in Brachypodium (Qin et al. 2017). To test the possibility that direct interaction between Hd3a and RCN regulates flowering, we examined the interaction between Hd3a and RCN3 in yeast. However, no direct interaction was observed (Supplementary Fig. S9C). Fig. 3 View largeDownload slide RCN competes with Hd3a for formation of a florigen activation complex. (A) Effect of RCN on OsMADS15 activation by the FAC. The amount of OsMADS15 RNA relative to that of Ubiquitin, when both Hd3a and OsFD1 were co-expressed, was set at 1. Values are means ± SEM from three independent experiments. Double asterisks denote a significant difference relative to OsMADS15 activation by the FAC without exogenous RCN expression as measured by t-test (P < 0.01). (B) Competitive binding of RCN and Hd3a to 14-3-3 examined by competitive in vitro pull-down assay. A mixture of Hd3a-HA and RCN3 was pulled down with GST–GF14c. Input ratios of Hd3a-HA to RCN3 (left) and RCN3 to Hd3a-HA (right) are indicated on the input gel images. Pull-down ratios of Hd3a-HA to RCN3 (left) and RCN3 to Hd3a-HA (right) are indicated below the pull-down gel images. (C) Co-localization of Hd3a and RCN3. BiFC assays between Hd3a and RCN3 (upper) and between Hd3a m2 and RCN3 m2 (lower) were performed. Hd3a m2, Hd3a R64K/R132K; RCN3 m2, RCN3 R62K/R130K. Scale bar = 5 μm. Fig. 3 View largeDownload slide RCN competes with Hd3a for formation of a florigen activation complex. (A) Effect of RCN on OsMADS15 activation by the FAC. The amount of OsMADS15 RNA relative to that of Ubiquitin, when both Hd3a and OsFD1 were co-expressed, was set at 1. Values are means ± SEM from three independent experiments. Double asterisks denote a significant difference relative to OsMADS15 activation by the FAC without exogenous RCN expression as measured by t-test (P < 0.01). (B) Competitive binding of RCN and Hd3a to 14-3-3 examined by competitive in vitro pull-down assay. A mixture of Hd3a-HA and RCN3 was pulled down with GST–GF14c. Input ratios of Hd3a-HA to RCN3 (left) and RCN3 to Hd3a-HA (right) are indicated on the input gel images. Pull-down ratios of Hd3a-HA to RCN3 (left) and RCN3 to Hd3a-HA (right) are indicated below the pull-down gel images. (C) Co-localization of Hd3a and RCN3. BiFC assays between Hd3a and RCN3 (upper) and between Hd3a m2 and RCN3 m2 (lower) were performed. Hd3a m2, Hd3a R64K/R132K; RCN3 m2, RCN3 R62K/R130K. Scale bar = 5 μm. RCN represses Hd3a activity to regulate inflorescence architecture in a 14-3-3-dependent manner It has been reported that ectopic expression of RCN delays flowering and increases branches in the inflorescence (Nakagawa et al. 2002, Wang et al. 2015). However, involvement of 14-3-3 in the regulation of reproductive development has not been analyzed. To address the issue, we made transgenic rice plants in which RCN3 and the m2 mutant are ectopically expressed under a strong constitutive promoter of maize Ubiquitin or RCN genes are simultaneously down-regulated by RNA interference (RNAi) (Miki et al. 2005). Consistent with previous reports, ectopic expression of RCN3 delayed flowering by 1 month and increased the number of branches and spikelets (Fig. 4). In contrast, ectopic expression of RCN3 m2 did not delay flowering (Fig. 4A) and did not significantly increase the number of spikelets (Fig. 4B), indicating the importance of 14-3-3 binding in the regulation of flowering time and inflorescence development by RCN3. In pUbq-RCN3 m2-GFP plants, the nuclear accumulation of RCN3–GFP was reduced compared with pUbq-RCN3-GFP plants (Supplementary Fig. S10B, C), supporting the proposal that FAC-like complex formation in the nucleus is impaired in RCN3 m2 (Fig. 2F, G). Fig. 4 View largeDownload slide Flowering time and inflorescence architecture of RCN transgenic plants. (A) Flowering time of transgenic plants under short-day conditions. (B) Inflorescence architecture of transgenic plants. (C) Inflorescence of transgenic plants. Values are means ± SEM. PB#, number of primary branches; SB#, number of secondary branches; SP#, number of spikelets; RCN3 m2, RCN3 R62K/R130K. Asterisks denote a significant difference relative to wild-type plants as measured by t-test (*P < 0.05, **P < 0.01). Scale bar = 2 cm. Fig. 4 View largeDownload slide Flowering time and inflorescence architecture of RCN transgenic plants. (A) Flowering time of transgenic plants under short-day conditions. (B) Inflorescence architecture of transgenic plants. (C) Inflorescence of transgenic plants. Values are means ± SEM. PB#, number of primary branches; SB#, number of secondary branches; SP#, number of spikelets; RCN3 m2, RCN3 R62K/R130K. Asterisks denote a significant difference relative to wild-type plants as measured by t-test (*P < 0.05, **P < 0.01). Scale bar = 2 cm. The expression pattern and the encoded amino acid sequence of the four RCN isoforms are very similar (Fig.1; Supplementary Fig. S1), and therefore the four RCN genes are thought to regulate plant development redundantly. We made triple (RCN1/RCN2/RCN3) and quadruple (RCN1/RCN2/RCN3/RCN4) RNAi plants and analyzed their reproductive development. The quadruple RCN RNAi plants showed slightly earlier flowering, whereas the triple RNAi plants flowered normally (Fig. 4A). Simultaneous knock down of the four RCN genes produces much smaller panicles with reduced branches (Liu et al. 2013), but phenotypical details of the reduction have not been analyzed. The triple and quadruple RCN RNAi plants had reduced numbers of spikelets and branches compared with wild-type plants (Fig. 4B, C). All the grains from quadruple RCN RNAi plants were unfilled and sterile (Supplementary Fig. S11C). A similar phenotype was observed in pUbq-Hd3a-GFP plants (Fig. 4). pUbq-Hd3a-GFP plants flowered a month earlier (Fig. 4A) and made fewer branches and spikelets than the wild type (Fig. 4B, C), although the degree of early flowering and reduction of branches and spikelets was much stronger than in quadruple RCN RNAi plants. All the grains from pUbq-Hd3a-GFP plants were also unfilled and sterile (Supplementary Fig. S11C). Discussion A model of how RCN antagonizes Hd3a We propose a model of how RCN antagonizes Hd3a to repress reproductive development (Fig. 5). Our data demonstrate that RCN as well as Hd3a can interact with OsFD1 through 14-3-3 binding and that increasing the amount of RCN can antagonize Hd3a in a 14-3-3-dependent manner. Here we named the proposed hexameric protein complex comprised of two RCNs, a 14-3-3 dimer and an OsFD1 dimer, as the florigen repression complex (FRC) (Fig. 5B). In our model, Hd3a competes with RCN for 14-3-3 binding, and the resulting balance between the FAC and FRC modulates florigen activity to optimize inflorescence development (Fig. 5C). The importance of balance between florigen and anti-florigen for plant development has been well studied in tomato (Shalit et al. 2009, Lifschitz et al. 2014). In tomato, SINGLE FLOWER TRUSS (SFT), a tomato FT homolog, regulates the reiterative growth of the sympodial unit, leaf maturation, and the complexity of compound leaves in a SELF PRUNING (SP)-, a tomato TFL1 homolog, dependent manner (Shalit et al. 2009). Our model provides the molecular basis to establish florigen/anti-florigen balance. Currently, the mode of complex formation does not appear to be different between FAC and FRC, except for the affinity for 14-3-3. Analysis of the crystal structure of FRC will reveal the molecular basis of the antagonistic function of FRC including different affinity for 14-3-3. Fig. 5 View largeDownload slide Model for the regulation of flowering and inflorescence development by competitive action of Hd3a and RCN in rice. (A) Illustration of the RCN expression pattern. RCN RNA is strongly and constitutively expressed in the phloem of the stem and beneath the SAM. Synthesized RCN protein is translocated to the SAM. (B) In the cells of the SAM, RCN competes with Hd3a for 14-3-3 binding and represses florigenic activity. RCN forms the florigen repression complex (FRC) with 14-3-3 and OsFD1, whereas Hd3a forms the FAC. The balance between the FRC and FAC is regulated depending on the ratio of Hd3a to RCN in the cell. (C) The FRC–FAC balance regulates the development of the SAM. At the vegetative phase, no Hd3a is accumulated in the SAM, and RCN makes the FRC. Under inductive SD conditions, Hd3a is produced and transported to the SAM. In the cells of the SAM, Hd3a competes with RCN for FAC formation. When the balance is shifted to the FAC, the reproductive program starts. In RCN knockdown plants, little FRC is formed. Once Hd3a reaches the cells of the SAM, it easily makes the FAC. The excess FAC impairs the optimal development of the inflorescence. Fig. 5 View largeDownload slide Model for the regulation of flowering and inflorescence development by competitive action of Hd3a and RCN in rice. (A) Illustration of the RCN expression pattern. RCN RNA is strongly and constitutively expressed in the phloem of the stem and beneath the SAM. Synthesized RCN protein is translocated to the SAM. (B) In the cells of the SAM, RCN competes with Hd3a for 14-3-3 binding and represses florigenic activity. RCN forms the florigen repression complex (FRC) with 14-3-3 and OsFD1, whereas Hd3a forms the FAC. The balance between the FRC and FAC is regulated depending on the ratio of Hd3a to RCN in the cell. (C) The FRC–FAC balance regulates the development of the SAM. At the vegetative phase, no Hd3a is accumulated in the SAM, and RCN makes the FRC. Under inductive SD conditions, Hd3a is produced and transported to the SAM. In the cells of the SAM, Hd3a competes with RCN for FAC formation. When the balance is shifted to the FAC, the reproductive program starts. In RCN knockdown plants, little FRC is formed. Once Hd3a reaches the cells of the SAM, it easily makes the FAC. The excess FAC impairs the optimal development of the inflorescence. A 14-3-3-independent mode of FT interaction and flowering regulation has been reported in Brachypodium (Qin et al. 2017). BdFT2b, an FT protein from an alternative splicing variant of BdFT2 in Brachypodium, cannot interact with FD or 14-3-3. However, it can form heterodimers with BdFT2a and FT1, thereby interfering with FAC formation to attenuate florigen activity (Qin et al. 2017). In addition, BdFT1 and BdFT2a, FT isoforms with an apparently typical FT structure, were also reported to make homo- and heterodimers in yeast (Qin et al. 2017). Our yeast-two hybrid experiment to test homo- and heterodimerization of Hd3a and RCN3 did not detect such a direct interaction. Considering that an alternative splicing variant of FT has not been discovered in rice (Qin et al. 2017), flowering regulation by direct FT dimerization would not exist in rice. However, it would be interesting to identify the structural features of FT/TFL1 that determine the efficiency of 14-3-3-independent dimerization. Regulation of reproductive development by balance between FAC and FRC is thought to be a general mechanism in plants because the 14-3-3-binding sites are highly conserved among the FT and TFL1 subfamilies of angiosperms (Taoka et al. 2011). According to our model, the amount and affinity for 14-3-3 of Hd3a and RCN are major determinants for the balance between the FAC and FRC. Our data indicate that RCN can bind 14-3-3 with five times higher affinity than Hd3a. The constitutive accumulation in SAM and higher affinity for 14-3-3 of RCN could be a barrier to protecting the SAM from inappropriately initiating the reproductive program with an insufficient amount of Hd3a. When a sufficient amount of Hd3a is transported into the cells of the SAM, the FRC would be converted to an FAC and the reproductive program would start. It remains to be elucidated how much Hd3a is necessary for conversion of the FRC to an FAC and initiation of the reproductive program. Quantification of trace amounts of Hd3a and RCN in the SAM is difficult due to current technical limitations. Imaging analysis of the localization of Hd3a and RCN in the same SAM will reveal the role of the local change of FAC–FRC balance in the regulation of reproductive development. The FAC or FT is thought to recruit an unidentified FT-interactor for transcriptional activation of the target genes (Ahn et al. 2006, Hanano and Goto 2011, Taoka et al. 2011, Ho and Weigel 2014), because FT does not possess any features typical of a transcriptional activator. Similarly, the repressive function of TFL1 may require an unidentified cofactor for transcriptional repression of the target genes. According to our model, competitive binding of TFL1 to 14-3-3 is sufficient to suppress FAC formation and, therefore, a TFL1-interacting cofactor is not a prerequisite. The facts that a few amino acid substitutions in FT can convert the floral promotor activity to a repressor activity (Hanzawa et al. 2005, Pin et al. 2010, Ho and Weigel 2014) and that the external loop region essential for functional specificity of the floral promoter/repressor of FT/TFL1 is not conserved in the TFL1 family (Ahn et al. 2006) further support this idea. RCN expression In some reports, RCN RNA has been detected in the shoot apex (Endo-Higashi and Izawa 2011, Liu et al. 2013). In contrast, RCN promoter activity in our gRCN-GUS plants was not detected in the SAM, from the vegetative to late reproductive phase (Fig. 5A). It might be possible that RCN RNA is transported from outside of the SAM because RNA of ATC, an Arabidopsis TFL1 homolog, moves a long distance (Huang et al. 2012). However, our RT–PCR analysis detected no accumulation of RCN RNA in the SAM. Considering that GUS signal was detected beneath the SAM in gRCN1-GUS, gRCN2-GUS and gRCN4-GUS plants, detection of RCN RNA in the SAM in the previous reports might be due to the contamination of RCN expression regions beneath the SAM during preparation of SAM tissue for RNA analysis. We carefully dissected the SAM for RNA extraction so that it did not contain the region beneath the SAM. Our data suggest that most RCN RNA is synthesized outside the SAM, although the possibility that a trace amount of RCN RNA below the detection limit of our semi-quantitative RT–PCR might be expressed or transported in the SAM cannot be excluded. The expression of RCN did not seem to be regulated by photoperiod, in contrast to Chrysanthemum AFT and Arabidopsis ATC, TFL1 homologs. AFT is induced in leaves under non-inductive LDs to repress FT activity (Higuchi et al. 2013). Similarly, Arabidopsis ATC is expressed under non-inductive SD conditions and contributes to repression of flowering (Huang et al. 2012). In sugar beet, BvFT1, a divergent FT paralog with three amino acid substitutions in the external loop, represses flowering to ensure perenniality (Pin et al. 2010). There are 13 FT homologs in rice, and only three among them (Hd3a, RFT1 and OsFTL1) have been extensively studied (Tsuji et al. 2013, Song et al. 2015). Like sugar beet, some of the rice FT homologs might have evolved to function as a floral repressor under non-inductive LD conditions. The expression and flowering function of these uncharacterized rice FT subfamilies remain to be analyzed. Currently, the reason why RCN should be expressed outside the SAM remains unknown. It could contribute to systemic co-ordination of the vegetative growth of the whole plant and reproductive development in SAM to produce an inflorescence with the optimal size and shape. Alternatively, RCN protein outside the SAM might serve to control florigen activity in tissues that express RCN, because Hd3a activates the downstream OsMADS14 and OsMADS15 genes in leaf, in addition to the SAM (Komiya et al. 2008). Transport of RCN protein In Arabidopsis, TFL1 protein moves locally in the shoot apex and maintains the meristem in an undifferentiated state (Conti and Bradley 2007). ATC RNA and the encoded protein can move from stock to scion and repress flowering (Huang et al. 2012). Our data demonstrated that RCN protein is translocated to the SAM. Proteome analysis of rice identified RCN3 peptide in the phloem sap (Aki et al. 2008), supporting our data on long-distance movement of RCN protein. As a component for long-distance transport of FT from phloem to shoot apex, FT-INTERACTING PROTEIN1 (FTIP1) and SODIUM POTASSIUMU ROOT DEFECTIVE 1 (NaKR1) have been identified (Liu et al. 2012, Zhu et al. 2016, Song et al. 2017). The overall protein structure of TFL1 is similar to that of FT, except for a few regions essential for floral promotion (Hanzawa et al. 2005, Ahn et al. 2006, Ho and Weigel 2014). If these FT transporters are also involved in the long-distance trafficking of TFL1 homologs, they could serve as a regulator between FT and TFL1. It remains to be determined whether these FT transporters also interact with TFL1 homologs. RCN functions in reproductive development Our data on RCN RNAi plants demonstrated that RCN does not seem to contribute strongly to determining the flowering time in rice. In contrast, the inflorescence architecture and the number of spikelets were greatly affected by knock down of RCN, consistent with data showing that RCN expression is not likely to be regulated by photoperiod, whereas Hd3a is up-regulated by SD conditions (Kojima et al. 2002). The slight early flowering of quadruple knockdown RCN plants could be explained by assuming that the threshold for Hd3a to make the FAC became lower due to the decreased RCN (Fig. 5B). In tomato, the function of SP, a tomato TFL1 homolog, depends on SFT, a tomato FT, whereby SP affects plant development as an anti-florigen by counteracting florigen SFT (Shalit et al. 2009). According to this model, it would be reasonable that a major phenotypic alteration in RCN RNAi plants appears after transition to the reproductive phase when florigen is up-regulated. In the RCN RNAi plants, most seeds were sterile and their grain yields were low. A similar, but stronger phenotype was observed in the plant that ectopically expressed Hd3a, in which secondary branches did not develop and a terminal flower was made. A major function of RCN is thought to be in maintaining the indeterminacy and undifferentiated state of the inflorescence meristem by repressing Hd3a activity at an appropriate level. A variety of genes and transposable elements are regulated by Hd3a in the SAM (Tamaki et al. 2015). Analyzing which developmental step is aborted in RCN-knockdown plants would be an important cue to dissect the florigen/anti-florigen function in reproductive development of rice. In conclusion, we presented evidence for FRC formation by RCN with 14-3-3 and OsFD1, and competition with Hd3a for FAC formation. RCN protein is translocated into the SAM and represses florigen activity at a level for optimal reproduction. Moderate alteration of the FAC can improve crop yield (Krieger et al. 2010, Park et al. 2014). Our work will contribute to fine-tuning the florigen pathway artificially for improvement of crop yields. Materials and Methods Plant materials and growth conditions Rice (Oryza sativa L. cv. Norin 8) was used as the wild type and parental cultivar for the transgenic study. Transgenic rice plants were generated using Agrobacterium-mediated transformation of rice calli (Hiei et al. 1994), and hygromycin-resistant plants were regenerated from transformed callus. Transgene expression was confirmed by semi-quantitative RT–PCR and observation of GFP fluorescence (Supplementary Figs. S10, S11). Plants were grown in climate chambers at 70% humidity under SD conditions with daily cycles of 10 h of light at 27.5°C and 14 h of dark at 25°C, or under LD conditions with daily cycles of 14 h of light at 27.5°C and 10 h of dark at 25°C. Flowering time was measured as days from seeding to heading. Transgenic plants in the T1 or T2 generation were used for analysis. For analysis of pUbq-Hd3a-GFP, RCN123 RNAi and RCN1234 RNAi plants, T0 plants were used due to their sterility. At least three independent lines were used for all phenotypic analysis. RNA extraction and semi-quantitative RT–PCR analysis The plant tissues were sampled for RNA extraction at Zeitgeber time (ZT) 0–4 from rice plants at vegetative [26–28 days post-imbibition (d.p.i.)], transition (30–31 d.p.i.), reproductive (40–42 d.p.i.) and late reproductive phase (49–50 d.p.i.). SAM was carefully sampled under a microscope to avoid leaf primordia and stem beneath the SAM. Total RNA was extracted using Trizol (Invitrogen) according to the manufacturer’s instructions. The cDNA was synthesized from 1 μg of RNA using oligo(dT) primer and Superscript II reverse transcriptase (Invitrogen). A 1 μl aliquot of cDNA was used for semi-quantitative RT–PCR of gene expression with gene-specific primers (Supplementary Table S1). Plasmid construction The coding regions and genomic regions of RCN were PCR-amplified with KOD FX Neo (TOYOBO) and cloned into the pENTER/D-TOPO (Invitrogen) to obtain entry clones. The Gateway destination vectors, pGWB, were used (Nakagawa et al. 2007). For constitutive expression, Cauliflower mosaic virus (CaMV) 35S promoter regions of pGWB vectors were substituted with the ubiquitin promoter of maize (pUbq) or the rolC promoter of A. rhizogenes (prolC). The p2KpANDA vector was used for RNAi trigger constructs (Miki and Shimamoto 2004). The nucleotide sequences of the plasmids were confirmed by sequence analysis. The oligonucleotides for PCR cloning are listed in Supplementary Table S1. The details are described in the Supplementary text. Histochemical GUS analysis GUS staining was performed as described previously (Tamaki et al. 2007). Samples were embedded in 4% low-melting agarose gel and sectioned at a thickness of 10–50 μm by using a VIBRATOME VIB3000plus (Meiwa). Sections were photographed with a BX50 microscope (Olympus) for observation of the SAM and vasculature, and with a MVX10 microscope (Olympus) for observation of the stem. Fluorescence imaging The SAM and stem sections of transgenic plants were observed with an LSM510 META (Carl Zeiss) as described previously (Tamaki et al. 2007). GFP signal was separated from background noise using an emission fingerprinting Liner Unmixing function. Protein expression and purification Preparation of recombinant protein was performed as in Taoka et al. (2011). The RCN genes were cloned into the pCold-GST vector (Hayashi and Kojima 2008) and expressed as GST fusion proteins in Escherichia coli BL21 Rosetta (DE3) (Novagen). The gene encoding GF14c was cloned into the pGEX 6 P-3 vector (GE Healthcare) and expressed as a GST-fused protein. The C-terminal 49 amino acid peptide of OsFD1 (147-195) with S192E phosphomimic substitution, which has been shown to be sufficient for 14-3-3 binding (Taoka et al., 2011), was expressed as a GST fusion protein from pCold-GST vector. The expressed GST-fused protein was purified by glutathione–Sepharose 4B resin (GE Healthcare). After digestion of the GST tag using GST–HRV 3 C protease, RCN proteins were purified by gel filtration chromatography using a Superdex75 column (GE Healthcare). The purified RCN protein was dialyzed against 50 mM potassium phosphate buffer (pH 6.8) containing 50 mM KCl, 1 mM dithiothreitol (DTT), 5 mM EDTA and 5 mM EGTA for 30 min at 4°C, and then against 50 mM potassium phosphate buffer (pH 6.8) containing 50 mM KCl and 1 mM DTT overnight at 4°C. GST pull-down assay GST pull-down assay was performed as in Taoka et al. (2011). Purified GST-tagged protein (10 nmol) was adsorbed onto 10 μl of glutathione–Sepharose 4B resin. For the negative control, 10 nmol of purified GST protein was used. After washout of excess GST-tagged proteins, 1 nmol of purified proteins lacking GST tags was added. After incubation at room temperature for 15 min, unbound proteins were washed away. Bound proteins were eluted together with GST-tagged proteins using 20 μl of phosphate buffer (pH 6.8) containing 50 mM KCl, 50 mM glutathione and 1 mM DTT. Eluted fractions were separated by SDS–PAGE and visualized by silver staining. The amount of bound protein was quantified by ImageJ software (NIH). For competitive pull-down assay, Hd3a protein tagged with the HA epitope at the C-terminus was prepared according to the same procedure as in the preparation of RCN proteins described above. Various ratios of Hd3a-HA to RCN3 protein were mixed with GST–GF14c protein and pulled down with glutathione–Sepharose 4B resin as described above. Bimolecular fluorescence complementation assay The RCN, GF14b, OsFD1 and GUS coding regions were cloned into fluorescent protein expression vectors or BiFC vectors and transformed into rice protoplasts (Taoka et al. 2011). The mCherry expression plasmid was introduced simultaneously as a marker for transformation efficiency. The nuclear localization signal (NLS)–cyan fluorescent protein (CFP) expression plasmid was introduced simultaneously as a marker for nuclear localization. To quantify protein–protein interactions from the BiFC assay, the fluorescence intensities of mVenus (BiFC) and mCherry (transformation marker) were measured under the same microscope settings, and the number of cells showing reliable BiFC and mCherry signals was scored (Taoka et al. 2011). Yeast two-hybrid assay pBTM116-GW and pVP16-GW were used as the Gateway destination vectors (Taoka et al. 2011). Transformation of yeast was performed using Frozen-EZ Yeast Transformation II™ (ZYMO RESEARCH) according to the manufacturer’s instruction. After transformation, yeast cells were grown at 30°C for 5 d on SC medium without uracil, tryptophan, leucine and histidine, containing added histidine or 1–10 mM 3-aminotriazole (3-AT) to test the bait–prey interaction. The concentration of 3-AT was determined by the bait–prey combination. Transient expression assay in rice protoplast Transformation of rice Oc protoplasts was performed as described previously (Taoka et al. 2011). For transient expression analysis, 4 μg of Hd3a expression vectors and 16 μg of OsFD1 expression vectors, together with or without 16 μg of RCN expression vectors were introduced into 500 μl of protoplast suspension at a concentration of 2 × 107 protoplasts ml–1 by the polyethylene glycol (PEG)-mediated transformation method. After a 24 h incubation at 30°C, the protoplast suspension was centrifuged and the cell pellet was frozen at –80°C for RNA extraction. cDNA (1 μl) was used for quantitative analysis of gene expression using SYBR Green PCR master mix (Applied Biosystems) with gene-specific primers. Data were collected using the ABI PRISM 7000 sequence detection system in accordance with the manufacturer’s instruction manual. Supplementary Data Supplementary data are available at PCP online. Funding This work was supported by the Japan Society for the Promotion of Science [KAKENHI grant Nos. 24000017 to K.S. and 17K07609 to K.T.] and by the Ministry of Education, Culture, Sports, Science and Technology [KAKENHI grant Nos. 26113512 to K.T., 16H06466 and 16H06464 to H.T.] Acknowledgments We thank Ms. Yuki Mistubayashi and Ms. Yoshiko Oono for technical assistance, Ms. Yuko Tamaki, Ms. Junko Naritomi, Ms. Yukiko Konomi and Mihoko Yoshino for making transgenic plants, and Ms. Momoko Yoneyama for biochemical assays. Disclosures The authors have no conflicts of interest to declare. References Abe M., Kobayashi Y., Yamamoto S., Daimon Y., Yamaguchi A., Ikeda Y., et al.   ( 2005) FD, a bZIP protein mediating signals from the floral pathway integrator FT at the shoot apex. Science  309: 1052– 1056. Google Scholar CrossRef Search ADS PubMed  Ahn J., Miller D., Winter V., Banfield M., Lee J., Yoo S., et al.   ( 2006) A divergent external loop confers antagonistic activity on floral regulators FT and TFL1. EMBO J.  25: 605– 614. Google Scholar CrossRef Search ADS PubMed  Aki T., Shigyo M., Nakano R., Yoneyama T., Yanagisawa S. ( 2008) Nano scale proteomics revealed the presence of regulatory proteins including three FT-Like proteins in phloem and xylem saps from rice. Plant Cell Physiol . 49: 767– 790. Google Scholar CrossRef Search ADS PubMed  Bradley D., Ratcliffe O., Vincent C., Carpenter R., Coen E. ( 1997) Inflorescence commitment and architecture in Arabidopsis. Science  275: 80– 83. Google Scholar CrossRef Search ADS PubMed  Conti L., Bradley D. ( 2007) TERMINAL FLOWER1 is a mobile signal controlling Arabidopsis architecture. Plant Cell  19: 767– 778. Google Scholar CrossRef Search ADS PubMed  Corbesier L., Vincent C., Jang S., Fornara F., Fan Q., Searle I, et al.   ( 2007) FT protein movement contributes to long-distance signaling in floral induction of Arabidopsis. Science  316: 1030– 1033. Google Scholar CrossRef Search ADS PubMed  de Boer A.H., van Kleeff P.J., Gao J. ( 2013) Plant 14-3-3 proteins as spiders in a web of phosphorylation. Protoplasma  250: 425– 440. Google Scholar CrossRef Search ADS PubMed  Endo-Higashi N., Izawa T. ( 2011) Flowering time genes Heading date 1 and Early heading date 1 together control panicle development in rice. Plant Cell Physiol . 52: 1083– 1094. Google Scholar CrossRef Search ADS PubMed  Hanano S., Goto K. ( 2011) Arabidopsis TERMINAL FLOWER1 is involved in the regulation of flowering time and inflorescence development through transcriptional repression. Plant Cell  23: 3172– 3184. Google Scholar CrossRef Search ADS PubMed  Hanzawa Y., Money T., Bradley D. ( 2005) A single amino acid converts a repressor to an activator of flowering. Proc. Natl. Acad. Sci. USA  102: 7748– 7753. Google Scholar CrossRef Search ADS   Hayama R., Yokoi S., Tamaki S., Yano M., Shimamoto K. ( 2003) Adaptation of photoperiodic control pathways produces short-day flowering in rice. Nature  422: 719– 722. Google Scholar CrossRef Search ADS PubMed  Hayashi K., Kojima C. ( 2008) pCold-GST vector: a novel cold-shock vector containing GST tag for soluble protein production. Protein Exp. Purif . 62: 120– 127. Google Scholar CrossRef Search ADS   Hiei Y., Ohta S., Komari T., Kumashiro T. ( 1994) Efficient transformation of rice (Oryza sativa L.) mediated by Agrobacterium and sequence analysis of the boundaries of the T-DNA. Plant J.  6: 271– 282. Google Scholar CrossRef Search ADS PubMed  Higuchi Y., Narumi T., Oda A., Nakano Y., Sumitomo K., Fukai S., et al.   ( 2013) The gated induction system of a systemic floral inhibitor, antiflorigen, determines obligate short-day flowering in chrysanthemums. Proc. Natl. Acad. Sci.USA  110: 17137– 17142. Google Scholar CrossRef Search ADS   Ho W., Weigel D. ( 2014) Structural features determining flower-promoting activity of Arabidopsis FLOWERING LOCUS T. Plant Cell  26: 552– 564. Google Scholar CrossRef Search ADS PubMed  Huang N., Jane W., Chen J., Yu T. ( 2012) Arabidopsis thaliana CENTRORADIALIS homologue (ATC) acts systemically to inhibit floral initiation in Arabidopsis. Plant J . 72: 175– 184. Google Scholar CrossRef Search ADS PubMed  Ishikawa R., Aoki M., Kurotani K., Yokoi S., Shinomura T., Takano M., et al.   ( 2011) Phytochrome B regulates Heading date 1 (Hd1)-mediated expression of rice florigen Hd3a and critical day length in rice. Mol. Genet. Genomics  285: 461– 470. Google Scholar CrossRef Search ADS PubMed  Jaeger K., Wigge P. ( 2007) FT protein acts as a long-range signal in Arabidopsis. Curr. Biol.  17: 1050– 1054. Google Scholar CrossRef Search ADS PubMed  Kardailsky I., Shukla V., Ahn J., Dagenais N., Christensen S., Nguyen J., et al.   ( 1999) Activation tagging of the floral inducer FT. Science  286: 1962– 1965. Google Scholar CrossRef Search ADS PubMed  Kawamoto N., Sasabe M., Endo M., Machida Y., Araki T. ( 2015) Calcium-dependent protein kinases responsible for the phosphorylation of a bZIP transcription factor FD crucial for the florigen complex formation. Sci. Rep.  5: 8341. Google Scholar CrossRef Search ADS PubMed  Kobayashi Y., Kaya H., Goto K., Iwabuchi M., Araki T. ( 1999) A pair of related genes with antagonistic roles in mediating flowering signals. Science  286: 1960– 1962. Google Scholar CrossRef Search ADS PubMed  Kobayashi K., Yasuno N., Sato Y., Yoda M., Yamazaki R., Kimizu M., et al.   ( 2012) Inflorescence meristem identity in rice is specified by overlapping functions of three AP1/FUL-like MADS box genes and PAP2, a SEPALLATA MADS box gene. Plant Cell  24: 1848– 1859. Google Scholar CrossRef Search ADS PubMed  Kojima S., Takahashi Y., Kobayashi Y., Monna L., Sasaki T., Araki T., et al.   ( 2002) Hd3a, a rice ortholog of the Arabidopsis FT gene, promotes transition to flowering downstream of Hd1 under short-day conditions. Plant Cell Physiol . 43: 1096– 1105. Google Scholar CrossRef Search ADS PubMed  Komiya R., Ikegami A., Tamaki S., Yokoi S., Shimamoto K. ( 2008) Hd3a and RFT1 are essential for flowering in rice. Development  135: 767– 774. Google Scholar CrossRef Search ADS PubMed  Krieger U., Lippman Z., Zamir D. ( 2010) The flowering gene SINGLE FLOWER TRUSS drives heterosis for yield in tomato. Nat. Genet.  42: 459– 463. Google Scholar CrossRef Search ADS PubMed  Lang A., Chailakhyan M., Frolova I. ( 1977) Promotion and inhibition of flower formation in a dayneutral plant in grafts with a short-day plant and a long-day plant. Proc. Natl. Acad. Sci. USA  74: 2412– 2416. Google Scholar CrossRef Search ADS   Lifschitz E., Ayre B., Eshed Y. ( 2014) Florigen and anti-florigen—a systemic mechanism for coordinating growth and termination in flowering plants. Front. Plant Sci.  5: 465. Google Scholar CrossRef Search ADS PubMed  Lin M., Belanger H., Lee Y., Varkonyi-Gasic E., Taoka K., Miura E., et al.   ( 2007) FLOWERING LOCUS T protein may act as the long-distance florigenic signal in the cucurbits. Plant Cell  19: 1488– 1506. Google Scholar CrossRef Search ADS PubMed  Liu L., Liu C., Hou X., Xi W., Shen L., Tao Z., et al.   ( 2012) FTIP1 is an essential regulator required for florigen transport. PLoS Biol.  10: e1001313. Google Scholar CrossRef Search ADS PubMed  Liu C., Teo Z., Bi Y., Song S., Xi W., Yang X., et al.   ( 2013) A conserved genetic pathway determines inflorescence architecture in Arabidopsis and rice. Dev. Cell  24: 612– 622. Google Scholar CrossRef Search ADS PubMed  Mathieu J., Warthmann N., Kuttner F., Schmid M. ( 2007) Export of FT protein from phloem companion cells is sufficient for floral induction in Arabidopsis. Curr. Biol . 17: 1055– 1060. Google Scholar CrossRef Search ADS PubMed  Miki D., Itoh R., Shimamoto K. ( 2005) RNA silencing of single and multiple members in a gene family of rice. Plant Physiol.  138: 1903– 1913. Google Scholar CrossRef Search ADS PubMed  Miki D., Shimamoto K. ( 2004) Simple RNAi vectors for stable and transient suppression of gene function in rice. Plant Cell Physiol.  45: 490– 495. Google Scholar CrossRef Search ADS PubMed  Mizoguchi T., Wright L., Fujiwara S., Cremer F., Lee K., Onouchi H., et al.   ( 2005) Distinct roles of GIGANTEA in promoting flowering and regulating circadian rhythms in Arabidopsis. Plant Cell  17: 2255– 2270. Google Scholar CrossRef Search ADS PubMed  Nakagawa T., Kurose T., Hino T., Tanaka K., Kawamukai M., Niwa Y., et al.   ( 2007) Development of series of gateway binary vectors, pGWBs, for realizing efficient construction of fusion genes for plant transformation. J. Biosci. Bioeng . 104: 34– 41. Google Scholar CrossRef Search ADS PubMed  Nakagawa M., Shimamoto K., Kyozuka J. ( 2002) Overexpression of RCN1 and RCN2, rice TERMINAL FLOWER 1/CENTRORADIALIS homologs, confers delay of phase transition and altered panicle morphology in rice. Plant J . 29: 743– 750. Google Scholar CrossRef Search ADS PubMed  Ormancey M., Thuleau P., Mazars C., Cotelle V. ( 2017) CDPKs and 14-3-3 proteins: emerging duo in signaling. Trends Plant Sci.  22: 263– 272. Google Scholar CrossRef Search ADS PubMed  Park S., Jiang K., Tal L., Yichie Y., Gar O., Zamir D., et al.   ( 2014) Optimization of crop productivity in tomato using induced mutations in the florigen pathway. Nat. Genet.  46: 1337– 1342. Google Scholar CrossRef Search ADS PubMed  Pin P., Benlloch R., Bonnet D., Wremerth-Weich E., Kraft T., Gielen J., et al.   ( 2010) An antagonistic pair of FT homologs mediates the control of flowering time in sugar beet. Science  330: 1397– 1400. Google Scholar CrossRef Search ADS PubMed  Qin Z., Wu J., Geng S., Feng N., Chen F., Kong X., et al.   ( 2017) Regulation of FT splicing by an endogenous cue in temperate grasses. Nat. Commun.  8: 14320. Google Scholar CrossRef Search ADS PubMed  Ryu J., Lee H., Seo P., Jung J., Ahn J., Park C. ( 2014) The Arabidopsis floral repressor BFT delays flowering by competing with FT for FD binding under high salinity. Mol. Plant  7: 377– 387. Google Scholar CrossRef Search ADS PubMed  Sawa M., Nusinow D., Kay S., Imaizumi T. ( 2007) FKF1 and GIGANTEA complex formation is required for day-length measurement in Arabidopsis. Science  318: 261– 265. Google Scholar CrossRef Search ADS PubMed  Shalit A., Rozman A., Goldshmidt A., Alvarez J., Bowman J., Eshed Y., et al.   ( 2009) The flowering hormone florigen functions as a general systemic regulator of growth and termination. Proc. Natl. Acad. Sci. USA  106: 8392– 8397. Google Scholar CrossRef Search ADS   Song S., Chen Y., Liu L., Wang Y., Bao S., Zhou X., et al.   ( 2017) OsFTIP1-mediated regulation of florigen transport in rice is negatively regulated by the ubiquitin-like domain kinase OsUbDKgamma4. Plant Cell  29: 491– 507. Google Scholar CrossRef Search ADS PubMed  Song Y., Shim J., Kinmonth-Schultz H., Imaizumi T. ( 2015) Photoperiodic flowering: time measurement mechanisms in leaves. Annu. Rev. Plant Biol.  66: 441– 464. Google Scholar CrossRef Search ADS PubMed  Suárez-López P., Wheatley K., Robson F., Onouchi H., Valverde F., Coupland G. ( 2001) CONSTANS mediates between the circadian clock and the control of flowering in Arabidopsis. Nature  410: 1116– 1120. Google Scholar CrossRef Search ADS PubMed  Tamaki S., Matsuo S., Wong H.L., Yokoi S., Shimamoto K. ( 2007) Hd3a protein is a mobile flowering signal in rice. Science  316: 1033– 1036. Google Scholar CrossRef Search ADS PubMed  Tamaki S., Tsuji H., Matsumoto A., Fujita A., Shimatani Z., Terada R., et al.   ( 2015) FT-like proteins induce transposon silencing in the shoot apex during floral induction in rice. Proc. Natl. Acad. Sci. USA  112: E901– E910. Google Scholar CrossRef Search ADS   Taoka K., Ohki I., Tsuji H., Furuita K., Hayashi K., Yanase T., et al.   ( 2011) 14-3-3 proteins act as intracellular receptors for rice Hd3a florigen. Nature  476: 332– 335. Google Scholar CrossRef Search ADS PubMed  Tsuji H., Taoka K., Shimamoto K. ( 2013) Florigen in rice: complex gene network for florigen transcription, florigen activation complex, and multiple functions. Curr. Opin. Plant Biol . 16: 228– 235. Google Scholar CrossRef Search ADS PubMed  Varkonyi-Gasic E., Moss S., Voogd C., Wang T., Putterill J., Hellens R. ( 2013) Homologs of FT, CEN and FD respond to developmental and environmental signals affecting growth and flowering in the perennial vine kiwifruit. New Phytol.  198: 732– 746. Google Scholar CrossRef Search ADS PubMed  Wang L., Sun S., Jin J., Fu D., Yang X., Weng X., et al.   ( 2015) Coordinated regulation of vegetative and reproductive branching in rice. Proc. Natl. Acad. Sci. USA  112: 15504– 15509. Google Scholar CrossRef Search ADS   Wigge P., Kim M., Jaeger K., Busch W., Schmid M., Lohmann J., et al.   ( 2005) Integration of spatial and temporal information during floral induction in Arabidopsis. Science  309: 1056– 1059. Google Scholar CrossRef Search ADS PubMed  Zhu Y., Liu L., Shen L., Yu H. ( 2016) NaKR1 regulates long-distance movement of FLOWERING LOCUS T in Arabidopsis. Nat. Plants  2: 16075. Google Scholar CrossRef Search ADS PubMed  Abbreviations Abbreviations AP1 APETALA1 3-AT 3-aminotriazole ATC Arabidopsis thaliana CENTRORADIALIS homolog BiFC bimolecular fluorescence complementation BMH1 brain modulosignalin homolog 1 CFP cyan fluorescent protein CO CONSTANS d.p.i. days post-imbibition DTT dithiothreitol FAC Florigen Activation Complex FRC Florigen Repression Complex FT FLOWERING LOCUS T GFP green fluorescent protein GI GIGANTEA GST glutathione S-transferase GUS β-glucuronidase HA hemagglutinin Hd1 Heading date 1 Hd3a Heading date 3a LD long day NLS nuclear localization signal PEBP phosphatidylethanolamine-binding protein RCN RICE CENTRORADIALIS RNAi RNA interference RT–PCR reverse transcription–PCR SAM shoot apical meristem SD short day SFT SINGLE FLOWER TRUSS SP SELF PRUNING TFL1 TERMINAL FLOWER 1 © The Author(s) 2018. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Plant and Cell Physiology Oxford University Press

TFL1-Like Proteins in Rice Antagonize Rice FT-Like Protein in Inflorescence Development by Competition for Complex Formation with 14-3-3 and FD

Loading next page...
 
/lp/ou_press/tfl1-like-proteins-in-rice-antagonize-rice-ft-like-protein-in-a3M4dH6aR2
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
0032-0781
eISSN
1471-9053
D.O.I.
10.1093/pcp/pcy021
Publisher site
See Article on Publisher Site

Abstract

Abstract Hd3a, a rice homolog of FLOWERING LOCUS T (FT), is a florigen that induces flowering. Hd3a forms a ternary ‘florigen activation complex’ (FAC) with 14-3-3 protein and OsFD1 transcription factor, a rice homolog of FD that induces transcription of OsMADS15, a rice homolog of APETALA1 (AP1), which leads to flowering. TERMINAL FLOWER 1 (TFL1) represses flowering and controls inflorescence architecture. However, the molecular basis for floral repression by TFL1 remains poorly understood. Here we show that RICE CENTRORADIALIS (RCN), rice TFL1-like proteins, compete with Hd3a for 14-3-3 binding. All four RCN genes are predominantly expressed in the vasculature, and RCN proteins are transported to the shoot apex to antagonize florigen activity and regulate inflorescence development. The antagonistic function of RCN to Hd3a is dependent on its 14-3-3 binding activity. Our results suggest a molecular basis for regulation of the balance between florigen FT and anti-florigen TFL1. Introduction Timing of flowering and inflorescence architecture are important for plant reproductive success and are controlled by both environmental and endogenous conditions. One of the most important environmental cues for flowering is photoperiod. Photoperiodic flowering is regulated by the interplay between light signaling and a circadian clock pathway (Song et al. 2015). In Arabidopsis, GIGANTEA (GI), a circadian clock-regulated nuclear protein, up-regulates CONSTANS (CO) (Mizoguchi et al. 2005, Sawa et al. 2007). CO is a CCT-domain transcription factor and activates the expression of FLOWERING LOCUS T (FT) (Suárez-López et al. 2001). FT and its homologs in other plant species act as the key protein component of florigen, a systemic floral inducer (Kardailsky et al. 1999, Kobayashi et al. 1999, Corbesier et al. 2007, Jaeger and Wigge 2007, Lin et al. 2007, Mathieu et al. 2007, Tamaki et al. 2007). The GI–CO–FT module for photoperiodic flowering is highly conserved in rice, a short-day (SD) plant. In rice, OsGI (rice ortholog of GI), Heading date 1 (Hd1; rice ortholog of CO) and Heading date 3a (Hd3a; a rice FT) make up the GI–CO–FT module (Kojima et al. 2002, Hayama et al. 2003, Tsuji et al. 2013). The dual function of Hd1 represents a major difference between Arabidopsis, a long-day (LD) plant, and rice; Hd1 activates and represses the expression of Hd3a under SD and LD conditions, respectively (Hayama et al. 2003, Ishikawa et al. 2011). FT encodes an approximately 20 kDa globular protein similar to phosphatidylethanolamine-binding protein (PEBP) (Kardailsky et al. 1999, Kobayashi et al. 1999). Loss of function of FT delays flowering and its ectopic expression causes early flowering and a terminal flower phenotype (Kardailsky et al. 1999, Kobayashi et al. 1999). FT protein is produced in the leaves under inductive day length conditions and transported through the phloem to the shoot apical meristem (SAM) to change the growth phase from vegetative to reproductive (Corbesier et al. 2007, Tamaki et al. 2007, Tamaki et al. 2015). The floral promotion activity of FT requires interaction with FD, a basic/leucine zipper (bZIP) transcription factor (Abe et al. 2005, Wigge et al. 2005). Phosphorylation of the conserved C-terminal motif of FD (SAP motif) is essential for interaction with FT and activation of APETALA1 (AP1), a floral meristem identity gene (Abe et al. 2005, Taoka et al. 2011, Kawamoto et al. 2015). In rice, it has been shown that the FT–FD interaction is mediated by 14-3-3 (Taoka et al. 2011). 14-3-3 is a highly conserved scaffold protein that bridges two proteins through phosphorylated 14-3-3-binding motifs and is involved in a variety of plant regulatory pathways (de Boer et al. 2013, Ormancey et al. 2017). Hd3a interacts with 14-3-3 in the cytoplasm and then the Hd3a–14-3-3 complex interacts with OsFD1, a rice FD homolog, in the nucleus through the phosphorylated SAP motif of OsFD1 (Taoka et al. 2011). Analysis of the crystal structure revealed that Hd3a makes a heterohexameric protein complex of two Hd3a, dimeric 14-3-3 and dimeric OsFD1, which is called the florigen activation complex (FAC) (Taoka et al. 2011). The amino acid residues critical for the interaction between FAC components are highly conserved among angiosperms, and mutations that impair the interaction reduce the activation of OsMADS15, a rice AP1 homolog (Taoka et al. 2011). TERMINAL FLOWER 1 (TFL1) is known to counteract and fine-tune florigen activity (Kobayashi et al. 1999, Shalit et al. 2009, Lifschitz et al. 2014). Loss of function of TFL1 causes early flowering and a terminal flower phenotype in Arabidopsis (Bradley et al. 1997). TFL1 also encodes a PEBP-like protein with similarity to FT (Bradley et al. 1997, Kardailsky et al. 1999, Kobayashi et al. 1999). However, small differences between TFL1 and FT in distribution of charged amino acid residues, structure of putative anion-binding pockets and the amino acid sequence of the external loop confer antagonistic activity to TFL1 (Hanzawa et al. 2005, Ahn et al. 2006, Ho and Weigel 2014). In the mature SAM, TFL1 mRNA is restricted to the center of the SAM, but TFL1 protein moves from cell to cell and co-ordinates the cell identity (Conti and Bradley 2007). In contrast, Arabidopsis thaliana CENTRORADIALIS homolog (ATC), another TFL1 in Arabidopsis, is expressed in the vasculature under SD conditions and is translocated to repress flowering (Huang et al. 2012). The graft-transmissible activity of floral inhibition suggests that ATC may be a systemic floral inhibitor, an anti-florigen that was proposed in tobacco (Lang et al. 1977, Huang et al. 2012). In rice, the four isoforms of RICE CENTRORADIALIS (RCN) are classified ino the TFL1 subfamily (Nakagawa et al. 2002, Liu et al. 2013). Overexpression of RCN1 and RCN2 delays flowering and increases the number of branches (Nakagawa et al. 2002). RCN-knockdown plants exhibit small panicles with reduced branches (Liu et al. 2013). TFL1 and TFL1-like proteins in several plant species are also known to interact with FD (Abe et al. 2005, Wigge et al. 2005, Hanano and Goto 2011, Huang et al. 2012, Higuchi et al. 2013, Varkonyi-Gasic et al. 2013, Ho and Weigel 2014, Ryu et al. 2014). The interaction with FD and conserved putative 14-3-3-binding sites in the TFL1 family suggest that TFL1 also makes a FAC-like complex to repress FT (Tsuji et al. 2013). However, the involvement of 14-3-3 in the antagonistic role of TFL1 has not been analyzed yet. In this study, we examined the spatiotemporal expression of RCN and the protein interaction of RCN. We show that RCN protein is transported to the SAM and can interact with OsFD1 to repress florigen activity in a 14-3-3-dependent manner. It suggests that competitive binding of RCN and Hd3a to 14-3-3 plays an important role in ensuring the optimal reproductive development of rice. Results Tissue-specific expression of RCN genes Phylogenetic tree analysis of 19 PEBP-like genes in rice identified four TFL1-like genes, RCN1–RCN4 (Supplementary Fig. S1). To examine the spatio-temporal expression of RCN RNA, semi-quantitative reverse transcription–PCR (RT–PCR) analysis was performed. RCN RNA was strongly detected in stem and root, from vegetative to late reproductive phase (Fig. 1A;Supplementary Fig. S2). In contrast, Hd3a is strongly expressed in the leaf under floral inductive SD conditions (Kojima et al. 2002, Tamaki et al. 2007). Under non-floral inductive LD conditions, RCN expression was not significantly different from that under SDs (Supplementary Fig. S2). In the SAM, no RCN RNA was detected at any growth phase (Fig. 1A). For a more sensitive assay, RT–PCR samples shown in Fig. 1A were subjected to a second round of PCR amplification, and consistently no amplification products were detected in the SAM (Supplementary Fig. S2A). To examine the expression pattern at the cellular level, the GUS (β-glucuronidase) reporter was fused with 5' upstream (approximately 3 kb) and coding regions of RCN genes (Supplementary Fig. S3A) and histochemical analysis of the transgenic rice plants with these chimeric genes (gRCN-GUS) was performed. GUS signal from gRCN1-GUS plants was detected in the vascular tissue of the stem and root (Fig. 1B;Supplementary Fig. S3). In the cross-section of the stem, GUS signal was localized in the phloem of the vasculature (Fig. 1B). Around the shoot apex, GUS signal was detected in the presumptive vascular tissue and beneath the SAM throughout the vegetative to reproductive phase (Fig. 1B). However, no GUS signal was detected in the SAM. Similar GUS staining patterns were observed in gRCN2-GUS and gRCN4-GUS plants (Fig. 1B;Supplementary Fig. S3). In gRCN3-GUS plants, GUS signal was also detected in the phloem, while it was not detected around the shoot apex (Fig. 1B). In all gRCN-GUS transgenic plants examined, no detectable GUS signal was found in the SAM from the vegetative to reproductive phase. Fig. 1 View largeDownload slide Expression of RCN in rice. (A) Expression of RCN RNA in rice. RCN expression was analyzed by semi-quantitative RT–PCR at the vegetative (V), transition (T), reproductive (R) and late reproductive (LR) phase. Ubiquitin (Ubiq) RNA was analyzed as a control. LB, leaf blade; LS, leaf sheath; ST, stem; RO, root; SAM, shoot apical meristem. (B) GUS staining images of gRCN-GUS transgenic rice plants. SAM at the vegetative (V) and reproductive (R) phase and the vasculature of the stem are shown. Asterisks indicate the SAM. A white triangle indicates GUS signal derived from the leaf primordium. Red triangles indicate phloem. Scale bar = 50 μm in the SAM and 20 μm in the vasculature. (C) Localization of RCN–GFP protein in gRCN-GFP transgenic rice plants. Confocal images of the SAM from the vegetative (V) to reproductive (R) phase and the stem vasculature in gRCN-GFP transgenic rice plants are shown. IM, inflorescence meristem; PBM, primary branch meristem; BR, bract. Scale bar = 50 μm. Fig. 1 View largeDownload slide Expression of RCN in rice. (A) Expression of RCN RNA in rice. RCN expression was analyzed by semi-quantitative RT–PCR at the vegetative (V), transition (T), reproductive (R) and late reproductive (LR) phase. Ubiquitin (Ubiq) RNA was analyzed as a control. LB, leaf blade; LS, leaf sheath; ST, stem; RO, root; SAM, shoot apical meristem. (B) GUS staining images of gRCN-GUS transgenic rice plants. SAM at the vegetative (V) and reproductive (R) phase and the vasculature of the stem are shown. Asterisks indicate the SAM. A white triangle indicates GUS signal derived from the leaf primordium. Red triangles indicate phloem. Scale bar = 50 μm in the SAM and 20 μm in the vasculature. (C) Localization of RCN–GFP protein in gRCN-GFP transgenic rice plants. Confocal images of the SAM from the vegetative (V) to reproductive (R) phase and the stem vasculature in gRCN-GFP transgenic rice plants are shown. IM, inflorescence meristem; PBM, primary branch meristem; BR, bract. Scale bar = 50 μm. RCN proteins are transported to the SAM Our RT–PCR and promoter–GUS analysis strongly suggested that no RCN RNA is expressed in the SAM. However, it has been reported that RCN regulates the floral transition and inflorescence development (Nakagawa et al. 2002, Liu et al. 2013, Wang et al. 2015). Therefore, we examined the localization of RCN proteins in plants. The same 5' upstream and coding regions of RCN genes as used in the GUS analysis were fused with the green fluorescent protein (GFP) coding region (gRCN-GFP) and fluorescence from RCN–GFP protein in the transgenic rice plants was observed with confocal microscopy (Fig. 1C;Supplementary Fig. S4). Consistent with the semi-quantitative RT–PCR and GUS staining, a strong RCN–GFP signal was observed in the stem vasculature for all four RCN isoforms (Fig. 1C). In addition, RCN–GFP signal was also observed in the SAM, from the vegetative to reproductive phase (Fig. 1C), indicating the transport activity of RCN proteins from the vascular tissue to the SAM. At the late reproductive phase, localization of RCN–GFP in the shoot apex was confined to the floral meristem for RCN1, RCN2 and RCN3, whereas RCN4–GFP was observed in and beneath the floral meristem (Supplementary Fig. S4). Translocation ability of RCN3–GFP to the SAM was also observed when RCN3–GFP was expressed under the control of a phloem-specific promoter from Agrobacterium rhizogenes, prolC (Supplementary Fig. S5). These results suggested that RCN proteins are predominantly synthesized in the phloem and translocated to the SAM. RCN can interact with 14-3-3 and OsFD1 First we tested the interaction of RCN with 14-3-3, because Hd3a interacts with OsFD1, a rice FD homolog, through 14-3-3 binding (Taoka et al. 2011). The critical amino acid residues of Hd3a for 14-3-3 binding were completely conserved among all four RCN isoforms (Fig. 2A). In a yeast two-hybrid assay, the four RCNs interacted with GF14b, a rice 14-3-3 (Fig. 2B). The 14-3-3 isoform preference of RCN was also comparable with that of Hd3a; the four RCNs interacted with GF14b, GF14c, GF14e and GF14f (Supplementary Fig. S6B). The amino acid substitution in RCN corresponding to impaired interaction between Hd3a and 14-3-3 reduced the interaction between RCN and 14-3-3 (Fig. 2B). The direct protein–protein interaction was tested by glutathione S-transferase (GST) pull-down assays using highly purified recombinant proteins. GST–GF14c pulled down the four recombinant RCN proteins with comparable efficiency (Fig. 2C). OsFD1 peptide was not pulled down with GST–RCN3, whereas GST–GF14c co-precipitated both RCN3 and OsFD1 (Fig. 2D). As with Hd3a, RCN3 was pulled down with GST–OsFD1 only in the presence of GF14c (Fig. 2D). The four RCNs interacted with OsFD1 in yeast (Supplementary Fig. S7B). Consistent with previous reports that phosphorylation of serine/threonine in the C-terminal SAP motif of Arabidopsis FD and rice OsFD1 is essential for 14-3-3 binding (Taoka et al. 2011, Kawamoto et al. 2015), alanine substitution of Ser192 (S192A) of OsFD1 prevented interaction with RCN, but phosphomimic glutamate substitution (S192E) retained it (Supplementary Fig. S6A). Endogenous yeast 14-3-3 was hypothesized to mediate the interaction between Hd3a and OsFD1, because Hd3a did not interact directly with OsFD1 in vitro (Taoka et al. 2011). In support of this idea, the two yeast 14-3-3s, brain modulosignalin homolog 1 (BMH1) and BMH2, interacted with RCN and Hd3a (Supplementary Fig. S6C) and OsFD1 (Supplementary Fig. S6D). These interactions were prevented with m2 substitution in the conserved 14-3-3 binding residues in RCN3 (R62K/R130K) and in Hd3a (R64K/R132K) or S192A substitution in OsFD1 (Supplementary Fig. S6). The RCN3–GF14b–OsFD1 interaction was also tested in rice protoplasts. Fluorescence from RCN3 tagged with mCherry was detected in both the nucleus and cytoplasm (Fig. 2E). In a bimolecular fluorescence complementation (BiFC) assay, fluorescence from RCN3–GF14b interaction was detected in the cytoplasm, and the fluorescence was significantly reduced by m2 substitution in RCN3 (Fig.2F;Supplementary Fig. S7D). BiFC fluorescence from RCN3–OsFD1 interaction was observed predominantly in the nucleus and m2 substitution also significantly reduced the interaction with OsFD1 (Fig. 2G;Supplementary Fig. S7E). The subcellular localization of RCN3 was not affected by m2 substitution (Supplementary Fig. S7A). These localization and interaction features were similar to those of RCN1 (Supplementary Fig. S7) and Hd3a (Taoka et al. 2011), suggesting that RCN interacts with OsFD1 through 14-3-3 binding, like Hd3a. Fig. 2 View largeDownload slide Interaction among RCN, 14-3-3 and OsFD1. (A) Schematic representation of RCN. The critical amino acid residues for Hd3a–14-3-3 interaction (R64, P96, F103 and R132) and their corresponding residues in RCN are shown. (B) Yeast two-hybrid assay between RCN and 14-3-3. 3-AT, 3-aminotriazole; His, histidine. (C) GST pull-down assay using GST–GF14c and RCN. The signal intensity relative to that of co-precipitated RCN3 is shown below. (D) GST pull-down assays using (from left to right) GST–RCN3 and OsFD1 peptide; GST–GF14c and RCN3; GST–GF14c and OsFD1; and GST–GF14c and RCN3 with or without GF14c. (E) Confocal images of cells expressing RCN3–mCherry. NLS–CFP was co-expressed as a nuclear marker protein. (F) BiFC assays showing interactions of RCN3–GF14b (left) and RCN3 m2–GF14b (right). Venus fluorescence in cells expressing the indicated proteins tagged with the N- or C-terminal halves of Venus is shown. mCherry was co-expressed as a marker for transformation and for nuclear and cytoplasmic localization. RCN3 m2: RCN3 R62K/R130K. Scale bar = 5 μm. (G) BiFC assays showing interactions of RCN3–OsFD1 (left) and RCN3 m2–OsFD1 (right). Scale bar = 5 μm. Fig. 2 View largeDownload slide Interaction among RCN, 14-3-3 and OsFD1. (A) Schematic representation of RCN. The critical amino acid residues for Hd3a–14-3-3 interaction (R64, P96, F103 and R132) and their corresponding residues in RCN are shown. (B) Yeast two-hybrid assay between RCN and 14-3-3. 3-AT, 3-aminotriazole; His, histidine. (C) GST pull-down assay using GST–GF14c and RCN. The signal intensity relative to that of co-precipitated RCN3 is shown below. (D) GST pull-down assays using (from left to right) GST–RCN3 and OsFD1 peptide; GST–GF14c and RCN3; GST–GF14c and OsFD1; and GST–GF14c and RCN3 with or without GF14c. (E) Confocal images of cells expressing RCN3–mCherry. NLS–CFP was co-expressed as a nuclear marker protein. (F) BiFC assays showing interactions of RCN3–GF14b (left) and RCN3 m2–GF14b (right). Venus fluorescence in cells expressing the indicated proteins tagged with the N- or C-terminal halves of Venus is shown. mCherry was co-expressed as a marker for transformation and for nuclear and cytoplasmic localization. RCN3 m2: RCN3 R62K/R130K. Scale bar = 5 μm. (G) BiFC assays showing interactions of RCN3–OsFD1 (left) and RCN3 m2–OsFD1 (right). Scale bar = 5 μm. RCN competes with Hd3a for 14-3-3 binding Interaction of RCN with 14-3-3 and OsFD1 led to a hypothesis that RCN antagonizes Hd3a through competition with 14-3-3 binding in formation of the FAC. To test this hypothesis, we first examined the relationship between floral repression and 14-3-3 binding activity of RCN. OsMADS15 is a floral meristem identity gene in rice that regulates the floral transition (Kobayashi et al. 2012) and is activated by Hd3a (Komiya et al. 2008). In rice protoplasts, co-expression of Hd3a and OsFD1 can activate the expression of OsMADS15 (Fig. 3A). However, the activation was significantly compromised when RCN was co-expressed, and the reduction of OsMADS15 was not observed when the RCN3 m2 mutant was co-expressed (Fig. 3A). This suggests that 14-3-3 binding of RCN plays an important role in acting against Hd3a. The competition between Hd3a and RCN for 14-3-3 binding was also analyzed by in vitro competitive pull-down assay with highly purified recombinant proteins. Various ratios of Hd3a–hemagglutinin (HA) to RCN3 protein were mixed, and Hd3a-HA and RCN were pulled down with GST–GF14c (Fig. 3B). The 50% inhibitory concentration (IC50) values were calculated from the signal intensity of the precipitated proteins and evaluated as 0.6 μM for RCN3 against Hd3a–GF14c binding and as 3.0 μM for Hd3a against RCN3–GF14c binding (Supplementary Fig. S8), indicating that RCN3 can bind to GF14c with approximately five times higher affinity than Hd3a in vitro. To assess if RCN can competitively form an FAC-like complex, we performed BiFC analyses between Hd3a and RCN. When RCN3 proteins fused with the N- and C-terminal half of mVenus were co-expressed with OsFD1 in rice protoplasts, fluorescence from RCN3-RCN3 interaction was observed in the nucleus (Supplementary Fig. S9A). The BiFC signal was significantly reduced when the m2 substitution was introduced in RCN3 (Supplementary Fig. S9A, B). The 14-3-3-dependent BiFC interaction of RCN3-RCN3 is consistent with a proposed structure of the FAC, a hexameric protein complex comprised of two Hd3a, a 14-3-3 dimer and an OsFD1 dimer (Taoka et al. 2011). When BiFC interaction of Hd3a-Hd3a was tested in the same way, 14-3-3-dependent BiFC interaction was observed (Supplementary Fig. S9A, B), supporting the notion that Hd3a makes a FAC. If RCN competes with 14-3-3 binding to Hd3a for FAC formation, a heteromeric FAC-like complex of Hd3a and RCN could be formed. As expected, 14-3-3-dependent BiFC interaction between Hd3a and RCN3/RCN1 was observed (Fig. 3C;Supplementary Fig. S9A, B), suggesting that Hd3a and RCN3 are exchangeable in formation of a FAC-like complex. Recently, direct interaction between FT isosforms was reported to fine-tune the flowering time in Brachypodium (Qin et al. 2017). To test the possibility that direct interaction between Hd3a and RCN regulates flowering, we examined the interaction between Hd3a and RCN3 in yeast. However, no direct interaction was observed (Supplementary Fig. S9C). Fig. 3 View largeDownload slide RCN competes with Hd3a for formation of a florigen activation complex. (A) Effect of RCN on OsMADS15 activation by the FAC. The amount of OsMADS15 RNA relative to that of Ubiquitin, when both Hd3a and OsFD1 were co-expressed, was set at 1. Values are means ± SEM from three independent experiments. Double asterisks denote a significant difference relative to OsMADS15 activation by the FAC without exogenous RCN expression as measured by t-test (P < 0.01). (B) Competitive binding of RCN and Hd3a to 14-3-3 examined by competitive in vitro pull-down assay. A mixture of Hd3a-HA and RCN3 was pulled down with GST–GF14c. Input ratios of Hd3a-HA to RCN3 (left) and RCN3 to Hd3a-HA (right) are indicated on the input gel images. Pull-down ratios of Hd3a-HA to RCN3 (left) and RCN3 to Hd3a-HA (right) are indicated below the pull-down gel images. (C) Co-localization of Hd3a and RCN3. BiFC assays between Hd3a and RCN3 (upper) and between Hd3a m2 and RCN3 m2 (lower) were performed. Hd3a m2, Hd3a R64K/R132K; RCN3 m2, RCN3 R62K/R130K. Scale bar = 5 μm. Fig. 3 View largeDownload slide RCN competes with Hd3a for formation of a florigen activation complex. (A) Effect of RCN on OsMADS15 activation by the FAC. The amount of OsMADS15 RNA relative to that of Ubiquitin, when both Hd3a and OsFD1 were co-expressed, was set at 1. Values are means ± SEM from three independent experiments. Double asterisks denote a significant difference relative to OsMADS15 activation by the FAC without exogenous RCN expression as measured by t-test (P < 0.01). (B) Competitive binding of RCN and Hd3a to 14-3-3 examined by competitive in vitro pull-down assay. A mixture of Hd3a-HA and RCN3 was pulled down with GST–GF14c. Input ratios of Hd3a-HA to RCN3 (left) and RCN3 to Hd3a-HA (right) are indicated on the input gel images. Pull-down ratios of Hd3a-HA to RCN3 (left) and RCN3 to Hd3a-HA (right) are indicated below the pull-down gel images. (C) Co-localization of Hd3a and RCN3. BiFC assays between Hd3a and RCN3 (upper) and between Hd3a m2 and RCN3 m2 (lower) were performed. Hd3a m2, Hd3a R64K/R132K; RCN3 m2, RCN3 R62K/R130K. Scale bar = 5 μm. RCN represses Hd3a activity to regulate inflorescence architecture in a 14-3-3-dependent manner It has been reported that ectopic expression of RCN delays flowering and increases branches in the inflorescence (Nakagawa et al. 2002, Wang et al. 2015). However, involvement of 14-3-3 in the regulation of reproductive development has not been analyzed. To address the issue, we made transgenic rice plants in which RCN3 and the m2 mutant are ectopically expressed under a strong constitutive promoter of maize Ubiquitin or RCN genes are simultaneously down-regulated by RNA interference (RNAi) (Miki et al. 2005). Consistent with previous reports, ectopic expression of RCN3 delayed flowering by 1 month and increased the number of branches and spikelets (Fig. 4). In contrast, ectopic expression of RCN3 m2 did not delay flowering (Fig. 4A) and did not significantly increase the number of spikelets (Fig. 4B), indicating the importance of 14-3-3 binding in the regulation of flowering time and inflorescence development by RCN3. In pUbq-RCN3 m2-GFP plants, the nuclear accumulation of RCN3–GFP was reduced compared with pUbq-RCN3-GFP plants (Supplementary Fig. S10B, C), supporting the proposal that FAC-like complex formation in the nucleus is impaired in RCN3 m2 (Fig. 2F, G). Fig. 4 View largeDownload slide Flowering time and inflorescence architecture of RCN transgenic plants. (A) Flowering time of transgenic plants under short-day conditions. (B) Inflorescence architecture of transgenic plants. (C) Inflorescence of transgenic plants. Values are means ± SEM. PB#, number of primary branches; SB#, number of secondary branches; SP#, number of spikelets; RCN3 m2, RCN3 R62K/R130K. Asterisks denote a significant difference relative to wild-type plants as measured by t-test (*P < 0.05, **P < 0.01). Scale bar = 2 cm. Fig. 4 View largeDownload slide Flowering time and inflorescence architecture of RCN transgenic plants. (A) Flowering time of transgenic plants under short-day conditions. (B) Inflorescence architecture of transgenic plants. (C) Inflorescence of transgenic plants. Values are means ± SEM. PB#, number of primary branches; SB#, number of secondary branches; SP#, number of spikelets; RCN3 m2, RCN3 R62K/R130K. Asterisks denote a significant difference relative to wild-type plants as measured by t-test (*P < 0.05, **P < 0.01). Scale bar = 2 cm. The expression pattern and the encoded amino acid sequence of the four RCN isoforms are very similar (Fig.1; Supplementary Fig. S1), and therefore the four RCN genes are thought to regulate plant development redundantly. We made triple (RCN1/RCN2/RCN3) and quadruple (RCN1/RCN2/RCN3/RCN4) RNAi plants and analyzed their reproductive development. The quadruple RCN RNAi plants showed slightly earlier flowering, whereas the triple RNAi plants flowered normally (Fig. 4A). Simultaneous knock down of the four RCN genes produces much smaller panicles with reduced branches (Liu et al. 2013), but phenotypical details of the reduction have not been analyzed. The triple and quadruple RCN RNAi plants had reduced numbers of spikelets and branches compared with wild-type plants (Fig. 4B, C). All the grains from quadruple RCN RNAi plants were unfilled and sterile (Supplementary Fig. S11C). A similar phenotype was observed in pUbq-Hd3a-GFP plants (Fig. 4). pUbq-Hd3a-GFP plants flowered a month earlier (Fig. 4A) and made fewer branches and spikelets than the wild type (Fig. 4B, C), although the degree of early flowering and reduction of branches and spikelets was much stronger than in quadruple RCN RNAi plants. All the grains from pUbq-Hd3a-GFP plants were also unfilled and sterile (Supplementary Fig. S11C). Discussion A model of how RCN antagonizes Hd3a We propose a model of how RCN antagonizes Hd3a to repress reproductive development (Fig. 5). Our data demonstrate that RCN as well as Hd3a can interact with OsFD1 through 14-3-3 binding and that increasing the amount of RCN can antagonize Hd3a in a 14-3-3-dependent manner. Here we named the proposed hexameric protein complex comprised of two RCNs, a 14-3-3 dimer and an OsFD1 dimer, as the florigen repression complex (FRC) (Fig. 5B). In our model, Hd3a competes with RCN for 14-3-3 binding, and the resulting balance between the FAC and FRC modulates florigen activity to optimize inflorescence development (Fig. 5C). The importance of balance between florigen and anti-florigen for plant development has been well studied in tomato (Shalit et al. 2009, Lifschitz et al. 2014). In tomato, SINGLE FLOWER TRUSS (SFT), a tomato FT homolog, regulates the reiterative growth of the sympodial unit, leaf maturation, and the complexity of compound leaves in a SELF PRUNING (SP)-, a tomato TFL1 homolog, dependent manner (Shalit et al. 2009). Our model provides the molecular basis to establish florigen/anti-florigen balance. Currently, the mode of complex formation does not appear to be different between FAC and FRC, except for the affinity for 14-3-3. Analysis of the crystal structure of FRC will reveal the molecular basis of the antagonistic function of FRC including different affinity for 14-3-3. Fig. 5 View largeDownload slide Model for the regulation of flowering and inflorescence development by competitive action of Hd3a and RCN in rice. (A) Illustration of the RCN expression pattern. RCN RNA is strongly and constitutively expressed in the phloem of the stem and beneath the SAM. Synthesized RCN protein is translocated to the SAM. (B) In the cells of the SAM, RCN competes with Hd3a for 14-3-3 binding and represses florigenic activity. RCN forms the florigen repression complex (FRC) with 14-3-3 and OsFD1, whereas Hd3a forms the FAC. The balance between the FRC and FAC is regulated depending on the ratio of Hd3a to RCN in the cell. (C) The FRC–FAC balance regulates the development of the SAM. At the vegetative phase, no Hd3a is accumulated in the SAM, and RCN makes the FRC. Under inductive SD conditions, Hd3a is produced and transported to the SAM. In the cells of the SAM, Hd3a competes with RCN for FAC formation. When the balance is shifted to the FAC, the reproductive program starts. In RCN knockdown plants, little FRC is formed. Once Hd3a reaches the cells of the SAM, it easily makes the FAC. The excess FAC impairs the optimal development of the inflorescence. Fig. 5 View largeDownload slide Model for the regulation of flowering and inflorescence development by competitive action of Hd3a and RCN in rice. (A) Illustration of the RCN expression pattern. RCN RNA is strongly and constitutively expressed in the phloem of the stem and beneath the SAM. Synthesized RCN protein is translocated to the SAM. (B) In the cells of the SAM, RCN competes with Hd3a for 14-3-3 binding and represses florigenic activity. RCN forms the florigen repression complex (FRC) with 14-3-3 and OsFD1, whereas Hd3a forms the FAC. The balance between the FRC and FAC is regulated depending on the ratio of Hd3a to RCN in the cell. (C) The FRC–FAC balance regulates the development of the SAM. At the vegetative phase, no Hd3a is accumulated in the SAM, and RCN makes the FRC. Under inductive SD conditions, Hd3a is produced and transported to the SAM. In the cells of the SAM, Hd3a competes with RCN for FAC formation. When the balance is shifted to the FAC, the reproductive program starts. In RCN knockdown plants, little FRC is formed. Once Hd3a reaches the cells of the SAM, it easily makes the FAC. The excess FAC impairs the optimal development of the inflorescence. A 14-3-3-independent mode of FT interaction and flowering regulation has been reported in Brachypodium (Qin et al. 2017). BdFT2b, an FT protein from an alternative splicing variant of BdFT2 in Brachypodium, cannot interact with FD or 14-3-3. However, it can form heterodimers with BdFT2a and FT1, thereby interfering with FAC formation to attenuate florigen activity (Qin et al. 2017). In addition, BdFT1 and BdFT2a, FT isoforms with an apparently typical FT structure, were also reported to make homo- and heterodimers in yeast (Qin et al. 2017). Our yeast-two hybrid experiment to test homo- and heterodimerization of Hd3a and RCN3 did not detect such a direct interaction. Considering that an alternative splicing variant of FT has not been discovered in rice (Qin et al. 2017), flowering regulation by direct FT dimerization would not exist in rice. However, it would be interesting to identify the structural features of FT/TFL1 that determine the efficiency of 14-3-3-independent dimerization. Regulation of reproductive development by balance between FAC and FRC is thought to be a general mechanism in plants because the 14-3-3-binding sites are highly conserved among the FT and TFL1 subfamilies of angiosperms (Taoka et al. 2011). According to our model, the amount and affinity for 14-3-3 of Hd3a and RCN are major determinants for the balance between the FAC and FRC. Our data indicate that RCN can bind 14-3-3 with five times higher affinity than Hd3a. The constitutive accumulation in SAM and higher affinity for 14-3-3 of RCN could be a barrier to protecting the SAM from inappropriately initiating the reproductive program with an insufficient amount of Hd3a. When a sufficient amount of Hd3a is transported into the cells of the SAM, the FRC would be converted to an FAC and the reproductive program would start. It remains to be elucidated how much Hd3a is necessary for conversion of the FRC to an FAC and initiation of the reproductive program. Quantification of trace amounts of Hd3a and RCN in the SAM is difficult due to current technical limitations. Imaging analysis of the localization of Hd3a and RCN in the same SAM will reveal the role of the local change of FAC–FRC balance in the regulation of reproductive development. The FAC or FT is thought to recruit an unidentified FT-interactor for transcriptional activation of the target genes (Ahn et al. 2006, Hanano and Goto 2011, Taoka et al. 2011, Ho and Weigel 2014), because FT does not possess any features typical of a transcriptional activator. Similarly, the repressive function of TFL1 may require an unidentified cofactor for transcriptional repression of the target genes. According to our model, competitive binding of TFL1 to 14-3-3 is sufficient to suppress FAC formation and, therefore, a TFL1-interacting cofactor is not a prerequisite. The facts that a few amino acid substitutions in FT can convert the floral promotor activity to a repressor activity (Hanzawa et al. 2005, Pin et al. 2010, Ho and Weigel 2014) and that the external loop region essential for functional specificity of the floral promoter/repressor of FT/TFL1 is not conserved in the TFL1 family (Ahn et al. 2006) further support this idea. RCN expression In some reports, RCN RNA has been detected in the shoot apex (Endo-Higashi and Izawa 2011, Liu et al. 2013). In contrast, RCN promoter activity in our gRCN-GUS plants was not detected in the SAM, from the vegetative to late reproductive phase (Fig. 5A). It might be possible that RCN RNA is transported from outside of the SAM because RNA of ATC, an Arabidopsis TFL1 homolog, moves a long distance (Huang et al. 2012). However, our RT–PCR analysis detected no accumulation of RCN RNA in the SAM. Considering that GUS signal was detected beneath the SAM in gRCN1-GUS, gRCN2-GUS and gRCN4-GUS plants, detection of RCN RNA in the SAM in the previous reports might be due to the contamination of RCN expression regions beneath the SAM during preparation of SAM tissue for RNA analysis. We carefully dissected the SAM for RNA extraction so that it did not contain the region beneath the SAM. Our data suggest that most RCN RNA is synthesized outside the SAM, although the possibility that a trace amount of RCN RNA below the detection limit of our semi-quantitative RT–PCR might be expressed or transported in the SAM cannot be excluded. The expression of RCN did not seem to be regulated by photoperiod, in contrast to Chrysanthemum AFT and Arabidopsis ATC, TFL1 homologs. AFT is induced in leaves under non-inductive LDs to repress FT activity (Higuchi et al. 2013). Similarly, Arabidopsis ATC is expressed under non-inductive SD conditions and contributes to repression of flowering (Huang et al. 2012). In sugar beet, BvFT1, a divergent FT paralog with three amino acid substitutions in the external loop, represses flowering to ensure perenniality (Pin et al. 2010). There are 13 FT homologs in rice, and only three among them (Hd3a, RFT1 and OsFTL1) have been extensively studied (Tsuji et al. 2013, Song et al. 2015). Like sugar beet, some of the rice FT homologs might have evolved to function as a floral repressor under non-inductive LD conditions. The expression and flowering function of these uncharacterized rice FT subfamilies remain to be analyzed. Currently, the reason why RCN should be expressed outside the SAM remains unknown. It could contribute to systemic co-ordination of the vegetative growth of the whole plant and reproductive development in SAM to produce an inflorescence with the optimal size and shape. Alternatively, RCN protein outside the SAM might serve to control florigen activity in tissues that express RCN, because Hd3a activates the downstream OsMADS14 and OsMADS15 genes in leaf, in addition to the SAM (Komiya et al. 2008). Transport of RCN protein In Arabidopsis, TFL1 protein moves locally in the shoot apex and maintains the meristem in an undifferentiated state (Conti and Bradley 2007). ATC RNA and the encoded protein can move from stock to scion and repress flowering (Huang et al. 2012). Our data demonstrated that RCN protein is translocated to the SAM. Proteome analysis of rice identified RCN3 peptide in the phloem sap (Aki et al. 2008), supporting our data on long-distance movement of RCN protein. As a component for long-distance transport of FT from phloem to shoot apex, FT-INTERACTING PROTEIN1 (FTIP1) and SODIUM POTASSIUMU ROOT DEFECTIVE 1 (NaKR1) have been identified (Liu et al. 2012, Zhu et al. 2016, Song et al. 2017). The overall protein structure of TFL1 is similar to that of FT, except for a few regions essential for floral promotion (Hanzawa et al. 2005, Ahn et al. 2006, Ho and Weigel 2014). If these FT transporters are also involved in the long-distance trafficking of TFL1 homologs, they could serve as a regulator between FT and TFL1. It remains to be determined whether these FT transporters also interact with TFL1 homologs. RCN functions in reproductive development Our data on RCN RNAi plants demonstrated that RCN does not seem to contribute strongly to determining the flowering time in rice. In contrast, the inflorescence architecture and the number of spikelets were greatly affected by knock down of RCN, consistent with data showing that RCN expression is not likely to be regulated by photoperiod, whereas Hd3a is up-regulated by SD conditions (Kojima et al. 2002). The slight early flowering of quadruple knockdown RCN plants could be explained by assuming that the threshold for Hd3a to make the FAC became lower due to the decreased RCN (Fig. 5B). In tomato, the function of SP, a tomato TFL1 homolog, depends on SFT, a tomato FT, whereby SP affects plant development as an anti-florigen by counteracting florigen SFT (Shalit et al. 2009). According to this model, it would be reasonable that a major phenotypic alteration in RCN RNAi plants appears after transition to the reproductive phase when florigen is up-regulated. In the RCN RNAi plants, most seeds were sterile and their grain yields were low. A similar, but stronger phenotype was observed in the plant that ectopically expressed Hd3a, in which secondary branches did not develop and a terminal flower was made. A major function of RCN is thought to be in maintaining the indeterminacy and undifferentiated state of the inflorescence meristem by repressing Hd3a activity at an appropriate level. A variety of genes and transposable elements are regulated by Hd3a in the SAM (Tamaki et al. 2015). Analyzing which developmental step is aborted in RCN-knockdown plants would be an important cue to dissect the florigen/anti-florigen function in reproductive development of rice. In conclusion, we presented evidence for FRC formation by RCN with 14-3-3 and OsFD1, and competition with Hd3a for FAC formation. RCN protein is translocated into the SAM and represses florigen activity at a level for optimal reproduction. Moderate alteration of the FAC can improve crop yield (Krieger et al. 2010, Park et al. 2014). Our work will contribute to fine-tuning the florigen pathway artificially for improvement of crop yields. Materials and Methods Plant materials and growth conditions Rice (Oryza sativa L. cv. Norin 8) was used as the wild type and parental cultivar for the transgenic study. Transgenic rice plants were generated using Agrobacterium-mediated transformation of rice calli (Hiei et al. 1994), and hygromycin-resistant plants were regenerated from transformed callus. Transgene expression was confirmed by semi-quantitative RT–PCR and observation of GFP fluorescence (Supplementary Figs. S10, S11). Plants were grown in climate chambers at 70% humidity under SD conditions with daily cycles of 10 h of light at 27.5°C and 14 h of dark at 25°C, or under LD conditions with daily cycles of 14 h of light at 27.5°C and 10 h of dark at 25°C. Flowering time was measured as days from seeding to heading. Transgenic plants in the T1 or T2 generation were used for analysis. For analysis of pUbq-Hd3a-GFP, RCN123 RNAi and RCN1234 RNAi plants, T0 plants were used due to their sterility. At least three independent lines were used for all phenotypic analysis. RNA extraction and semi-quantitative RT–PCR analysis The plant tissues were sampled for RNA extraction at Zeitgeber time (ZT) 0–4 from rice plants at vegetative [26–28 days post-imbibition (d.p.i.)], transition (30–31 d.p.i.), reproductive (40–42 d.p.i.) and late reproductive phase (49–50 d.p.i.). SAM was carefully sampled under a microscope to avoid leaf primordia and stem beneath the SAM. Total RNA was extracted using Trizol (Invitrogen) according to the manufacturer’s instructions. The cDNA was synthesized from 1 μg of RNA using oligo(dT) primer and Superscript II reverse transcriptase (Invitrogen). A 1 μl aliquot of cDNA was used for semi-quantitative RT–PCR of gene expression with gene-specific primers (Supplementary Table S1). Plasmid construction The coding regions and genomic regions of RCN were PCR-amplified with KOD FX Neo (TOYOBO) and cloned into the pENTER/D-TOPO (Invitrogen) to obtain entry clones. The Gateway destination vectors, pGWB, were used (Nakagawa et al. 2007). For constitutive expression, Cauliflower mosaic virus (CaMV) 35S promoter regions of pGWB vectors were substituted with the ubiquitin promoter of maize (pUbq) or the rolC promoter of A. rhizogenes (prolC). The p2KpANDA vector was used for RNAi trigger constructs (Miki and Shimamoto 2004). The nucleotide sequences of the plasmids were confirmed by sequence analysis. The oligonucleotides for PCR cloning are listed in Supplementary Table S1. The details are described in the Supplementary text. Histochemical GUS analysis GUS staining was performed as described previously (Tamaki et al. 2007). Samples were embedded in 4% low-melting agarose gel and sectioned at a thickness of 10–50 μm by using a VIBRATOME VIB3000plus (Meiwa). Sections were photographed with a BX50 microscope (Olympus) for observation of the SAM and vasculature, and with a MVX10 microscope (Olympus) for observation of the stem. Fluorescence imaging The SAM and stem sections of transgenic plants were observed with an LSM510 META (Carl Zeiss) as described previously (Tamaki et al. 2007). GFP signal was separated from background noise using an emission fingerprinting Liner Unmixing function. Protein expression and purification Preparation of recombinant protein was performed as in Taoka et al. (2011). The RCN genes were cloned into the pCold-GST vector (Hayashi and Kojima 2008) and expressed as GST fusion proteins in Escherichia coli BL21 Rosetta (DE3) (Novagen). The gene encoding GF14c was cloned into the pGEX 6 P-3 vector (GE Healthcare) and expressed as a GST-fused protein. The C-terminal 49 amino acid peptide of OsFD1 (147-195) with S192E phosphomimic substitution, which has been shown to be sufficient for 14-3-3 binding (Taoka et al., 2011), was expressed as a GST fusion protein from pCold-GST vector. The expressed GST-fused protein was purified by glutathione–Sepharose 4B resin (GE Healthcare). After digestion of the GST tag using GST–HRV 3 C protease, RCN proteins were purified by gel filtration chromatography using a Superdex75 column (GE Healthcare). The purified RCN protein was dialyzed against 50 mM potassium phosphate buffer (pH 6.8) containing 50 mM KCl, 1 mM dithiothreitol (DTT), 5 mM EDTA and 5 mM EGTA for 30 min at 4°C, and then against 50 mM potassium phosphate buffer (pH 6.8) containing 50 mM KCl and 1 mM DTT overnight at 4°C. GST pull-down assay GST pull-down assay was performed as in Taoka et al. (2011). Purified GST-tagged protein (10 nmol) was adsorbed onto 10 μl of glutathione–Sepharose 4B resin. For the negative control, 10 nmol of purified GST protein was used. After washout of excess GST-tagged proteins, 1 nmol of purified proteins lacking GST tags was added. After incubation at room temperature for 15 min, unbound proteins were washed away. Bound proteins were eluted together with GST-tagged proteins using 20 μl of phosphate buffer (pH 6.8) containing 50 mM KCl, 50 mM glutathione and 1 mM DTT. Eluted fractions were separated by SDS–PAGE and visualized by silver staining. The amount of bound protein was quantified by ImageJ software (NIH). For competitive pull-down assay, Hd3a protein tagged with the HA epitope at the C-terminus was prepared according to the same procedure as in the preparation of RCN proteins described above. Various ratios of Hd3a-HA to RCN3 protein were mixed with GST–GF14c protein and pulled down with glutathione–Sepharose 4B resin as described above. Bimolecular fluorescence complementation assay The RCN, GF14b, OsFD1 and GUS coding regions were cloned into fluorescent protein expression vectors or BiFC vectors and transformed into rice protoplasts (Taoka et al. 2011). The mCherry expression plasmid was introduced simultaneously as a marker for transformation efficiency. The nuclear localization signal (NLS)–cyan fluorescent protein (CFP) expression plasmid was introduced simultaneously as a marker for nuclear localization. To quantify protein–protein interactions from the BiFC assay, the fluorescence intensities of mVenus (BiFC) and mCherry (transformation marker) were measured under the same microscope settings, and the number of cells showing reliable BiFC and mCherry signals was scored (Taoka et al. 2011). Yeast two-hybrid assay pBTM116-GW and pVP16-GW were used as the Gateway destination vectors (Taoka et al. 2011). Transformation of yeast was performed using Frozen-EZ Yeast Transformation II™ (ZYMO RESEARCH) according to the manufacturer’s instruction. After transformation, yeast cells were grown at 30°C for 5 d on SC medium without uracil, tryptophan, leucine and histidine, containing added histidine or 1–10 mM 3-aminotriazole (3-AT) to test the bait–prey interaction. The concentration of 3-AT was determined by the bait–prey combination. Transient expression assay in rice protoplast Transformation of rice Oc protoplasts was performed as described previously (Taoka et al. 2011). For transient expression analysis, 4 μg of Hd3a expression vectors and 16 μg of OsFD1 expression vectors, together with or without 16 μg of RCN expression vectors were introduced into 500 μl of protoplast suspension at a concentration of 2 × 107 protoplasts ml–1 by the polyethylene glycol (PEG)-mediated transformation method. After a 24 h incubation at 30°C, the protoplast suspension was centrifuged and the cell pellet was frozen at –80°C for RNA extraction. cDNA (1 μl) was used for quantitative analysis of gene expression using SYBR Green PCR master mix (Applied Biosystems) with gene-specific primers. Data were collected using the ABI PRISM 7000 sequence detection system in accordance with the manufacturer’s instruction manual. Supplementary Data Supplementary data are available at PCP online. Funding This work was supported by the Japan Society for the Promotion of Science [KAKENHI grant Nos. 24000017 to K.S. and 17K07609 to K.T.] and by the Ministry of Education, Culture, Sports, Science and Technology [KAKENHI grant Nos. 26113512 to K.T., 16H06466 and 16H06464 to H.T.] Acknowledgments We thank Ms. Yuki Mistubayashi and Ms. Yoshiko Oono for technical assistance, Ms. Yuko Tamaki, Ms. Junko Naritomi, Ms. Yukiko Konomi and Mihoko Yoshino for making transgenic plants, and Ms. Momoko Yoneyama for biochemical assays. Disclosures The authors have no conflicts of interest to declare. References Abe M., Kobayashi Y., Yamamoto S., Daimon Y., Yamaguchi A., Ikeda Y., et al.   ( 2005) FD, a bZIP protein mediating signals from the floral pathway integrator FT at the shoot apex. Science  309: 1052– 1056. Google Scholar CrossRef Search ADS PubMed  Ahn J., Miller D., Winter V., Banfield M., Lee J., Yoo S., et al.   ( 2006) A divergent external loop confers antagonistic activity on floral regulators FT and TFL1. EMBO J.  25: 605– 614. Google Scholar CrossRef Search ADS PubMed  Aki T., Shigyo M., Nakano R., Yoneyama T., Yanagisawa S. ( 2008) Nano scale proteomics revealed the presence of regulatory proteins including three FT-Like proteins in phloem and xylem saps from rice. Plant Cell Physiol . 49: 767– 790. Google Scholar CrossRef Search ADS PubMed  Bradley D., Ratcliffe O., Vincent C., Carpenter R., Coen E. ( 1997) Inflorescence commitment and architecture in Arabidopsis. Science  275: 80– 83. Google Scholar CrossRef Search ADS PubMed  Conti L., Bradley D. ( 2007) TERMINAL FLOWER1 is a mobile signal controlling Arabidopsis architecture. Plant Cell  19: 767– 778. Google Scholar CrossRef Search ADS PubMed  Corbesier L., Vincent C., Jang S., Fornara F., Fan Q., Searle I, et al.   ( 2007) FT protein movement contributes to long-distance signaling in floral induction of Arabidopsis. Science  316: 1030– 1033. Google Scholar CrossRef Search ADS PubMed  de Boer A.H., van Kleeff P.J., Gao J. ( 2013) Plant 14-3-3 proteins as spiders in a web of phosphorylation. Protoplasma  250: 425– 440. Google Scholar CrossRef Search ADS PubMed  Endo-Higashi N., Izawa T. ( 2011) Flowering time genes Heading date 1 and Early heading date 1 together control panicle development in rice. Plant Cell Physiol . 52: 1083– 1094. Google Scholar CrossRef Search ADS PubMed  Hanano S., Goto K. ( 2011) Arabidopsis TERMINAL FLOWER1 is involved in the regulation of flowering time and inflorescence development through transcriptional repression. Plant Cell  23: 3172– 3184. Google Scholar CrossRef Search ADS PubMed  Hanzawa Y., Money T., Bradley D. ( 2005) A single amino acid converts a repressor to an activator of flowering. Proc. Natl. Acad. Sci. USA  102: 7748– 7753. Google Scholar CrossRef Search ADS   Hayama R., Yokoi S., Tamaki S., Yano M., Shimamoto K. ( 2003) Adaptation of photoperiodic control pathways produces short-day flowering in rice. Nature  422: 719– 722. Google Scholar CrossRef Search ADS PubMed  Hayashi K., Kojima C. ( 2008) pCold-GST vector: a novel cold-shock vector containing GST tag for soluble protein production. Protein Exp. Purif . 62: 120– 127. Google Scholar CrossRef Search ADS   Hiei Y., Ohta S., Komari T., Kumashiro T. ( 1994) Efficient transformation of rice (Oryza sativa L.) mediated by Agrobacterium and sequence analysis of the boundaries of the T-DNA. Plant J.  6: 271– 282. Google Scholar CrossRef Search ADS PubMed  Higuchi Y., Narumi T., Oda A., Nakano Y., Sumitomo K., Fukai S., et al.   ( 2013) The gated induction system of a systemic floral inhibitor, antiflorigen, determines obligate short-day flowering in chrysanthemums. Proc. Natl. Acad. Sci.USA  110: 17137– 17142. Google Scholar CrossRef Search ADS   Ho W., Weigel D. ( 2014) Structural features determining flower-promoting activity of Arabidopsis FLOWERING LOCUS T. Plant Cell  26: 552– 564. Google Scholar CrossRef Search ADS PubMed  Huang N., Jane W., Chen J., Yu T. ( 2012) Arabidopsis thaliana CENTRORADIALIS homologue (ATC) acts systemically to inhibit floral initiation in Arabidopsis. Plant J . 72: 175– 184. Google Scholar CrossRef Search ADS PubMed  Ishikawa R., Aoki M., Kurotani K., Yokoi S., Shinomura T., Takano M., et al.   ( 2011) Phytochrome B regulates Heading date 1 (Hd1)-mediated expression of rice florigen Hd3a and critical day length in rice. Mol. Genet. Genomics  285: 461– 470. Google Scholar CrossRef Search ADS PubMed  Jaeger K., Wigge P. ( 2007) FT protein acts as a long-range signal in Arabidopsis. Curr. Biol.  17: 1050– 1054. Google Scholar CrossRef Search ADS PubMed  Kardailsky I., Shukla V., Ahn J., Dagenais N., Christensen S., Nguyen J., et al.   ( 1999) Activation tagging of the floral inducer FT. Science  286: 1962– 1965. Google Scholar CrossRef Search ADS PubMed  Kawamoto N., Sasabe M., Endo M., Machida Y., Araki T. ( 2015) Calcium-dependent protein kinases responsible for the phosphorylation of a bZIP transcription factor FD crucial for the florigen complex formation. Sci. Rep.  5: 8341. Google Scholar CrossRef Search ADS PubMed  Kobayashi Y., Kaya H., Goto K., Iwabuchi M., Araki T. ( 1999) A pair of related genes with antagonistic roles in mediating flowering signals. Science  286: 1960– 1962. Google Scholar CrossRef Search ADS PubMed  Kobayashi K., Yasuno N., Sato Y., Yoda M., Yamazaki R., Kimizu M., et al.   ( 2012) Inflorescence meristem identity in rice is specified by overlapping functions of three AP1/FUL-like MADS box genes and PAP2, a SEPALLATA MADS box gene. Plant Cell  24: 1848– 1859. Google Scholar CrossRef Search ADS PubMed  Kojima S., Takahashi Y., Kobayashi Y., Monna L., Sasaki T., Araki T., et al.   ( 2002) Hd3a, a rice ortholog of the Arabidopsis FT gene, promotes transition to flowering downstream of Hd1 under short-day conditions. Plant Cell Physiol . 43: 1096– 1105. Google Scholar CrossRef Search ADS PubMed  Komiya R., Ikegami A., Tamaki S., Yokoi S., Shimamoto K. ( 2008) Hd3a and RFT1 are essential for flowering in rice. Development  135: 767– 774. Google Scholar CrossRef Search ADS PubMed  Krieger U., Lippman Z., Zamir D. ( 2010) The flowering gene SINGLE FLOWER TRUSS drives heterosis for yield in tomato. Nat. Genet.  42: 459– 463. Google Scholar CrossRef Search ADS PubMed  Lang A., Chailakhyan M., Frolova I. ( 1977) Promotion and inhibition of flower formation in a dayneutral plant in grafts with a short-day plant and a long-day plant. Proc. Natl. Acad. Sci. USA  74: 2412– 2416. Google Scholar CrossRef Search ADS   Lifschitz E., Ayre B., Eshed Y. ( 2014) Florigen and anti-florigen—a systemic mechanism for coordinating growth and termination in flowering plants. Front. Plant Sci.  5: 465. Google Scholar CrossRef Search ADS PubMed  Lin M., Belanger H., Lee Y., Varkonyi-Gasic E., Taoka K., Miura E., et al.   ( 2007) FLOWERING LOCUS T protein may act as the long-distance florigenic signal in the cucurbits. Plant Cell  19: 1488– 1506. Google Scholar CrossRef Search ADS PubMed  Liu L., Liu C., Hou X., Xi W., Shen L., Tao Z., et al.   ( 2012) FTIP1 is an essential regulator required for florigen transport. PLoS Biol.  10: e1001313. Google Scholar CrossRef Search ADS PubMed  Liu C., Teo Z., Bi Y., Song S., Xi W., Yang X., et al.   ( 2013) A conserved genetic pathway determines inflorescence architecture in Arabidopsis and rice. Dev. Cell  24: 612– 622. Google Scholar CrossRef Search ADS PubMed  Mathieu J., Warthmann N., Kuttner F., Schmid M. ( 2007) Export of FT protein from phloem companion cells is sufficient for floral induction in Arabidopsis. Curr. Biol . 17: 1055– 1060. Google Scholar CrossRef Search ADS PubMed  Miki D., Itoh R., Shimamoto K. ( 2005) RNA silencing of single and multiple members in a gene family of rice. Plant Physiol.  138: 1903– 1913. Google Scholar CrossRef Search ADS PubMed  Miki D., Shimamoto K. ( 2004) Simple RNAi vectors for stable and transient suppression of gene function in rice. Plant Cell Physiol.  45: 490– 495. Google Scholar CrossRef Search ADS PubMed  Mizoguchi T., Wright L., Fujiwara S., Cremer F., Lee K., Onouchi H., et al.   ( 2005) Distinct roles of GIGANTEA in promoting flowering and regulating circadian rhythms in Arabidopsis. Plant Cell  17: 2255– 2270. Google Scholar CrossRef Search ADS PubMed  Nakagawa T., Kurose T., Hino T., Tanaka K., Kawamukai M., Niwa Y., et al.   ( 2007) Development of series of gateway binary vectors, pGWBs, for realizing efficient construction of fusion genes for plant transformation. J. Biosci. Bioeng . 104: 34– 41. Google Scholar CrossRef Search ADS PubMed  Nakagawa M., Shimamoto K., Kyozuka J. ( 2002) Overexpression of RCN1 and RCN2, rice TERMINAL FLOWER 1/CENTRORADIALIS homologs, confers delay of phase transition and altered panicle morphology in rice. Plant J . 29: 743– 750. Google Scholar CrossRef Search ADS PubMed  Ormancey M., Thuleau P., Mazars C., Cotelle V. ( 2017) CDPKs and 14-3-3 proteins: emerging duo in signaling. Trends Plant Sci.  22: 263– 272. Google Scholar CrossRef Search ADS PubMed  Park S., Jiang K., Tal L., Yichie Y., Gar O., Zamir D., et al.   ( 2014) Optimization of crop productivity in tomato using induced mutations in the florigen pathway. Nat. Genet.  46: 1337– 1342. Google Scholar CrossRef Search ADS PubMed  Pin P., Benlloch R., Bonnet D., Wremerth-Weich E., Kraft T., Gielen J., et al.   ( 2010) An antagonistic pair of FT homologs mediates the control of flowering time in sugar beet. Science  330: 1397– 1400. Google Scholar CrossRef Search ADS PubMed  Qin Z., Wu J., Geng S., Feng N., Chen F., Kong X., et al.   ( 2017) Regulation of FT splicing by an endogenous cue in temperate grasses. Nat. Commun.  8: 14320. Google Scholar CrossRef Search ADS PubMed  Ryu J., Lee H., Seo P., Jung J., Ahn J., Park C. ( 2014) The Arabidopsis floral repressor BFT delays flowering by competing with FT for FD binding under high salinity. Mol. Plant  7: 377– 387. Google Scholar CrossRef Search ADS PubMed  Sawa M., Nusinow D., Kay S., Imaizumi T. ( 2007) FKF1 and GIGANTEA complex formation is required for day-length measurement in Arabidopsis. Science  318: 261– 265. Google Scholar CrossRef Search ADS PubMed  Shalit A., Rozman A., Goldshmidt A., Alvarez J., Bowman J., Eshed Y., et al.   ( 2009) The flowering hormone florigen functions as a general systemic regulator of growth and termination. Proc. Natl. Acad. Sci. USA  106: 8392– 8397. Google Scholar CrossRef Search ADS   Song S., Chen Y., Liu L., Wang Y., Bao S., Zhou X., et al.   ( 2017) OsFTIP1-mediated regulation of florigen transport in rice is negatively regulated by the ubiquitin-like domain kinase OsUbDKgamma4. Plant Cell  29: 491– 507. Google Scholar CrossRef Search ADS PubMed  Song Y., Shim J., Kinmonth-Schultz H., Imaizumi T. ( 2015) Photoperiodic flowering: time measurement mechanisms in leaves. Annu. Rev. Plant Biol.  66: 441– 464. Google Scholar CrossRef Search ADS PubMed  Suárez-López P., Wheatley K., Robson F., Onouchi H., Valverde F., Coupland G. ( 2001) CONSTANS mediates between the circadian clock and the control of flowering in Arabidopsis. Nature  410: 1116– 1120. Google Scholar CrossRef Search ADS PubMed  Tamaki S., Matsuo S., Wong H.L., Yokoi S., Shimamoto K. ( 2007) Hd3a protein is a mobile flowering signal in rice. Science  316: 1033– 1036. Google Scholar CrossRef Search ADS PubMed  Tamaki S., Tsuji H., Matsumoto A., Fujita A., Shimatani Z., Terada R., et al.   ( 2015) FT-like proteins induce transposon silencing in the shoot apex during floral induction in rice. Proc. Natl. Acad. Sci. USA  112: E901– E910. Google Scholar CrossRef Search ADS   Taoka K., Ohki I., Tsuji H., Furuita K., Hayashi K., Yanase T., et al.   ( 2011) 14-3-3 proteins act as intracellular receptors for rice Hd3a florigen. Nature  476: 332– 335. Google Scholar CrossRef Search ADS PubMed  Tsuji H., Taoka K., Shimamoto K. ( 2013) Florigen in rice: complex gene network for florigen transcription, florigen activation complex, and multiple functions. Curr. Opin. Plant Biol . 16: 228– 235. Google Scholar CrossRef Search ADS PubMed  Varkonyi-Gasic E., Moss S., Voogd C., Wang T., Putterill J., Hellens R. ( 2013) Homologs of FT, CEN and FD respond to developmental and environmental signals affecting growth and flowering in the perennial vine kiwifruit. New Phytol.  198: 732– 746. Google Scholar CrossRef Search ADS PubMed  Wang L., Sun S., Jin J., Fu D., Yang X., Weng X., et al.   ( 2015) Coordinated regulation of vegetative and reproductive branching in rice. Proc. Natl. Acad. Sci. USA  112: 15504– 15509. Google Scholar CrossRef Search ADS   Wigge P., Kim M., Jaeger K., Busch W., Schmid M., Lohmann J., et al.   ( 2005) Integration of spatial and temporal information during floral induction in Arabidopsis. Science  309: 1056– 1059. Google Scholar CrossRef Search ADS PubMed  Zhu Y., Liu L., Shen L., Yu H. ( 2016) NaKR1 regulates long-distance movement of FLOWERING LOCUS T in Arabidopsis. Nat. Plants  2: 16075. Google Scholar CrossRef Search ADS PubMed  Abbreviations Abbreviations AP1 APETALA1 3-AT 3-aminotriazole ATC Arabidopsis thaliana CENTRORADIALIS homolog BiFC bimolecular fluorescence complementation BMH1 brain modulosignalin homolog 1 CFP cyan fluorescent protein CO CONSTANS d.p.i. days post-imbibition DTT dithiothreitol FAC Florigen Activation Complex FRC Florigen Repression Complex FT FLOWERING LOCUS T GFP green fluorescent protein GI GIGANTEA GST glutathione S-transferase GUS β-glucuronidase HA hemagglutinin Hd1 Heading date 1 Hd3a Heading date 3a LD long day NLS nuclear localization signal PEBP phosphatidylethanolamine-binding protein RCN RICE CENTRORADIALIS RNAi RNA interference RT–PCR reverse transcription–PCR SAM shoot apical meristem SD short day SFT SINGLE FLOWER TRUSS SP SELF PRUNING TFL1 TERMINAL FLOWER 1 © The Author(s) 2018. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oup.com

Journal

Plant and Cell PhysiologyOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off